Two data scientists say they have created AI algorithms that can do in a week what biological researchers might otherwise spend years trying to pull off in a laboratory: discover antibody-based treatments that have a fighting chance to beat back COVID-19.

In fact, studies have shown it takes an average of five years and half a billion dollars to find and fine-tune antibodies in a lab, Andrew Satz and Brett Averso, both execs of a 12-member startup called EVQLV, explain.

Speaking with their alma mater, Columbia Universitys Data Science Institute, Satz and Averso say their machine-learning algorithms can help by cutting the chances of costly experimental failures in the lab.

We fail in the computer as much as possible to reduce the possibility of downstream failure in the laboratory, Satz tells the institutes news division. [T]hat shaves a significant amount of time from laborious and time-consuming work.

See the original post:
Machine learning teams with antibody science on COVID-19 treatment discovery - AI in Healthcare

Machine learning (ML) is a type of artificial intelligence (AI) that allows software applications to become more accurate at predicting outcomes without being explicitly programmed to do so. Machine learning algorithms use historical data as input to predict new output values.

Recommendation enginesare a common use case for machine learning. Other popular uses include fraud detection, spam filtering, malware threat detection, business process automation (BPA) andpredictive maintenance.

Classical machine learning is often categorized by how an algorithm learns to become more accurate in its predictions. There are two basic approaches: supervised learning andunsupervised learning. The type of algorithm a data scientist chooses to use is dependent upon what type of data they want to predict.

Supervised machine learning requires thedata scientist to train the algorithm with both labeled inputs and desired outputs. Supervised learning algorithms are good for the following tasks:

Unsupervised ML algorithms do not require data to be labeled. They sift through unlabeled data to look for patterns that can be used to group data points into subsets.Unsupervised learning algorithms are good for the following tasks:

Today, machine learning is used in a wide range of applications. Perhaps one of the most well-known examples of machine learning in action is the recommendation engine that powers Facebook's News Feed.

Facebook uses machine learning to personalize how each member's feed is delivered. If a member frequently stops to read a particular groups posts, the recommendation engine will start to show more of that groups activity earlier in the feed.

Behind the scenes, the engine is attempting to reinforce known patterns in the members online behavior. Should the member change patterns and fail to read posts from that group in the coming weeks, the News Feed will adjust accordingly.

In addition to recommendation engines, other uses for machine learning include the following:

Customer relationship management CRM software can use machine learning models to analyze email and prompt sales team members to respond to the most important messages first. More advanced systems can even recommend potentially effective responses.

Business intelligence- BI and analytics vendors use machine learning in their software to identify potentially important data points, patterns of data points and anomalies.

Human resource information systems HRIS systems can use machine learning models to filter through applications and identify the best candidates for an open position.

Self-driving cars Machine learning algorithms can even make it possible for a semi-autonomous car to recognize a partially visible object and alert the driver.

Virtual assistants- Smart assistants typically combine supervised and unsupervised machine learning models to interpret natural speech and supply context.

The process of choosing the right machine learning model to solve a problem can be time consuming if not approached strategically.

Step 1: Align the problem with potential data inputs that should be considered for the solution. This step requires help from data scientists and experts who have a deep understanding of the problem.

Step 2: Collect data, format it and label the data if necessary. This step is typically led by data scientists, with help from data wranglers.

Step 3: Chose which algorithm(s) to use and >test to see how well they perform. This step is usually carried out by data scientists.

Step 4: Continue to fine tune outputs until they reach an acceptable level of accuracy. This step is usually carried out by data scientists with feedback from experts who have a deep understanding of the problem.

Explaining how a specific ML model works can be challenging when the model is complex. There are some vertical industries where data scientists have to use simple machine learning models because its important for the business to explain how each and every decision was made. This is especially true in industries with heavy compliance burdens like banking and insurance.

Complex models can accurate predictions, but explaining to a lay person how an output was determined can be difficult.

While machine learning algorithms have been around for decades, they've attained new popularity asartificial intelligence(AI) has grown in prominence. Deep learning models, in particular, powers today's most advanced AI applications.

Machine learning platforms are among enterprise technology's most competitive realms, with most major vendors, including Amazon, Google, Microsoft, IBM and others, racing to sign customers up for platform services that cover the spectrum of machine learning activities, including data collection,data preparation, data classification, model building, training and application deployment.

As machine learning continues to increase in importance to business operations and AI becomes ever more practical in enterprise settings, the machine learning platform wars will only intensify.

Continued research into deep learning and AI is increasingly focused on developing more general applications. Today's AI models require extensive training in order to produce an algorithm that is highly optimized to perform one task. But some researchers are exploring ways to make models more flexible and are seeking techniques that allow a machine to apply context learned from one task to future, different tasks.

1642 - Blaise Pascal invents a mechanical machine that can add, subtract, multiply and divide.

1679 - Gottfried Wilhelm Leibniz devises the system of binary code.

1834 - Charles Babbage conceives the idea for a general all-purpose device that could be programmed with punched cards.

1842 - Ada Lovelace describes a sequence of operations for solving mathematical problems using Charles' Babbage's theoretical punch-card machine and becomes the first programmer.

1847 - George Boole creates Boolean logic, a form of algebra in which all values can be reduced to the binary values of true or false.

1936 - English logician and cryptanalyst AlanTuring proposes a Universal Machine that could decipher and execute a set of instructions. His published proof is considered the basis of computer science.

1952 - Arthur Samuel creates a program to help an IBM computer get better at checkers the more it plays.

1959 - MADALINE becomes the first artificial neural network applied to a real-world problem: removing echoes from phone lines.

1985 - Terry Sejnowski and Charles Rosenbergs artificial neural network taught itself how to correctly pronounce 20,000 words in one week.

1997 - IBMs Deep Blue beat chess grandmaster Garry Kasparov.

1999 - A CAD prototype intelligent workstation reviewed 22,000 mammograms and detected cancer 52% more accurately than radiologists did.

2006 - Computer scientist Geoffrey Hinton invents the term deep learning to describe neural net research.

2012 - An unsupervised neural network created by Google learned to recognize cats in YouTube videos with 74.8% accuracy.

2014 - A chatbot passes the Turing Test by convincing 33% of human judges that it was a Ukrainian teen named Eugene Goostman.

2014 - Googles AlphaGo defeats the human champion in Go, the most difficult board game in the world.

2016 - LipNet, DeepMinds artificial-intelligence system, identifies lip-read words in video with an accuracy of 93.4%.

2019 - Amazon controls 70% of the market share for virtual assistants in the U.S.

Originally posted here:
What is machine learning (ML)? - Definition from WhatIs.com

What is Machine Learning?

Machine Learning is a system that can learn from example through self-improvement and without being explicitly coded by programmer. The breakthrough comes with the idea that a machine can singularly learn from the data (i.e., example) to produce accurate results.

Machine learning combines data with statistical tools to predict an output. This output is then used by corporate to makes actionable insights. Machine learning is closely related to data mining and Bayesian predictive modeling. The machine receives data as input, use an algorithm to formulate answers.

A typical machine learning tasks are to provide a recommendation. For those who have a Netflix account, all recommendations of movies or series are based on the user's historical data. Tech companies are using unsupervised learning to improve the user experience with personalizing recommendation.

Machine learning is also used for a variety of task like fraud detection, predictive maintenance, portfolio optimization, automatize task and so on.

In this basic tutorial, you will learn-

Traditional programming differs significantly from machine learning. In traditional programming, a programmer code all the rules in consultation with an expert in the industry for which software is being developed. Each rule is based on a logical foundation; the machine will execute an output following the logical statement. When the system grows complex, more rules need to be written. It can quickly become unsustainable to maintain.

Machine learning is supposed to overcome this issue. The machine learns how the input and output data are correlated and it writes a rule. The programmers do not need to write new rules each time there is new data. The algorithms adapt in response to new data and experiences to improve efficacy over time.

Machine learning is the brain where all the learning takes place. The way the machine learns is similar to the human being. Humans learn from experience. The more we know, the more easily we can predict. By analogy, when we face an unknown situation, the likelihood of success is lower than the known situation. Machines are trained the same. To make an accurate prediction, the machine sees an example. When we give the machine a similar example, it can figure out the outcome. However, like a human, if its feed a previously unseen example, the machine has difficulties to predict.

The core objective of machine learning is the learning and inference. First of all, the machine learns through the discovery of patterns. This discovery is made thanks to the data. One crucial part of the data scientist is to choose carefully which data to provide to the machine. The list of attributes used to solve a problem is called a feature vector. You can think of a feature vector as a subset of data that is used to tackle a problem.

The machine uses some fancy algorithms to simplify the reality and transform this discovery into a model. Therefore, the learning stage is used to describe the data and summarize it into a model.

For instance, the machine is trying to understand the relationship between the wage of an individual and the likelihood to go to a fancy restaurant. It turns out the machine finds a positive relationship between wage and going to a high-end restaurant: This is the model

When the model is built, it is possible to test how powerful it is on never-seen-before data. The new data are transformed into a features vector, go through the model and give a prediction. This is all the beautiful part of machine learning. There is no need to update the rules or train again the model. You can use the model previously trained to make inference on new data.

The life of Machine Learning programs is straightforward and can be summarized in the following points:

Once the algorithm gets good at drawing the right conclusions, it applies that knowledge to new sets of data.

Machine learning can be grouped into two broad learning tasks: Supervised and Unsupervised. There are many other algorithms

An algorithm uses training data and feedback from humans to learn the relationship of given inputs to a given output. For instance, a practitioner can use marketing expense and weather forecast as input data to predict the sales of cans.

You can use supervised learning when the output data is known. The algorithm will predict new data.

There are two categories of supervised learning:

Imagine you want to predict the gender of a customer for a commercial. You will start gathering data on the height, weight, job, salary, purchasing basket, etc. from your customer database. You know the gender of each of your customer, it can only be male or female. The objective of the classifier will be to assign a probability of being a male or a female (i.e., the label) based on the information (i.e., features you have collected). When the model learned how to recognize male or female, you can use new data to make a prediction. For instance, you just got new information from an unknown customer, and you want to know if it is a male or female. If the classifier predicts male = 70%, it means the algorithm is sure at 70% that this customer is a male, and 30% it is a female.

The label can be of two or more classes. The above example has only two classes, but if a classifier needs to predict object, it has dozens of classes (e.g., glass, table, shoes, etc. each object represents a class)

When the output is a continuous value, the task is a regression. For instance, a financial analyst may need to forecast the value of a stock based on a range of feature like equity, previous stock performances, macroeconomics index. The system will be trained to estimate the price of the stocks with the lowest possible error.

In unsupervised learning, an algorithm explores input data without being given an explicit output variable (e.g., explores customer demographic data to identify patterns)

You can use it when you do not know how to classify the data, and you want the algorithm to find patterns and classify the data for you

Type

K-means clustering

Puts data into some groups (k) that each contains data with similar characteristics (as determined by the model, not in advance by humans)

Clustering

Gaussian mixture model

A generalization of k-means clustering that provides more flexibility in the size and shape of groups (clusters

Clustering

Hierarchical clustering

Splits clusters along a hierarchical tree to form a classification system.

Can be used for Cluster loyalty-card customer

Clustering

Recommender system

Help to define the relevant data for making a recommendation.

Clustering

PCA/T-SNE

Mostly used to decrease the dimensionality of the data. The algorithms reduce the number of features to 3 or 4 vectors with the highest variances.

Dimension Reduction

There are plenty of machine learning algorithms. The choice of the algorithm is based on the objective.

In the example below, the task is to predict the type of flower among the three varieties. The predictions are based on the length and the width of the petal. The picture depicts the results of ten different algorithms. The picture on the top left is the dataset. The data is classified into three categories: red, light blue and dark blue. There are some groupings. For instance, from the second image, everything in the upper left belongs to the red category, in the middle part, there is a mixture of uncertainty and light blue while the bottom corresponds to the dark category. The other images show different algorithms and how they try to classified the data.

The primary challenge of machine learning is the lack of data or the diversity in the dataset. A machine cannot learn if there is no data available. Besides, a dataset with a lack of diversity gives the machine a hard time. A machine needs to have heterogeneity to learn meaningful insight. It is rare that an algorithm can extract information when there are no or few variations. It is recommended to have at least 20 observations per group to help the machine learn. This constraint leads to poor evaluation and prediction.

Augmentation:

Automation:

Finance Industry

Government organization

Healthcare industry

Marketing

Example of application of Machine Learning in Supply Chain

Machine learning gives terrific results for visual pattern recognition, opening up many potential applications in physical inspection and maintenance across the entire supply chain network.

Unsupervised learning can quickly search for comparable patterns in the diverse dataset. In turn, the machine can perform quality inspection throughout the logistics hub, shipment with damage and wear.

For instance, IBM's Watson platform can determine shipping container damage. Watson combines visual and systems-based data to track, report and make recommendations in real-time.

In past year stock manager relies extensively on the primary method to evaluate and forecast the inventory. When combining big data and machine learning, better forecasting techniques have been implemented (an improvement of 20 to 30 % over traditional forecasting tools). In term of sales, it means an increase of 2 to 3 % due to the potential reduction in inventory costs.

Example of Machine Learning Google Car

For example, everybody knows the Google car. The car is full of lasers on the roof which are telling it where it is regarding the surrounding area. It has radar in the front, which is informing the car of the speed and motion of all the cars around it. It uses all of that data to figure out not only how to drive the car but also to figure out and predict what potential drivers around the car are going to do. What's impressive is that the car is processing almost a gigabyte a second of data.

Machine learning is the best tool so far to analyze, understand and identify a pattern in the data. One of the main ideas behind machine learning is that the computer can be trained to automate tasks that would be exhaustive or impossible for a human being. The clear breach from the traditional analysis is that machine learning can take decisions with minimal human intervention.

Take the following example; a retail agent can estimate the price of a house based on his own experience and his knowledge of the market.

A machine can be trained to translate the knowledge of an expert into features. The features are all the characteristics of a house, neighborhood, economic environment, etc. that make the price difference. For the expert, it took him probably some years to master the art of estimate the price of a house. His expertise is getting better and better after each sale.

For the machine, it takes millions of data, (i.e., example) to master this art. At the very beginning of its learning, the machine makes a mistake, somehow like the junior salesman. Once the machine sees all the example, it got enough knowledge to make its estimation. At the same time, with incredible accuracy. The machine is also able to adjust its mistake accordingly.

Most of the big company have understood the value of machine learning and holding data. McKinsey have estimated that the value of analytics ranges from $9.5 trillion to $15.4 trillion while $5 to 7 trillion can be attributed to the most advanced AI techniques.

Go here to see the original:
Machine Learning Tutorial for Beginners

AUSTIN, Texas, March 23, 2020 /PRNewswire-PRWeb/ --Novi Labs ("Novi") today announced the release of Novi Prediction Engine version 2.0. This provides critical economic data to E&P workflows such as well planning or acquisition & divestitures. Novi customers can now run a wide range of large-scale scenarios in minutes and get immediate feedback on the economic feasibility of each plan. As price headwinds face the industry, having the ability to quickly and easily evaluate hundreds of scenarios allows operators to efficiently allocate capital.

In addition to the economic outputs, Novi Prediction Engine 2.0 also includes new features targeting enhanced usability and increased efficiency. Novi is now publishing confidence intervals as a standard output for every prediction. This allows customers to understand how confident the model is of each prediction it makes, which is critical decision-making criterion. A video demonstration of Novi Prediction Engine version 2.0 is available at https://novilabs.com/prediction-engine-v2/.

"With the integration of economic outputs and confidence intervals into Novi Prediction Engine, customers have increased leverage, transparency and certainty in what the Novi models are providing in support of their business decisions. This form of rapid scenario driven testing that is unlocked by the Novi platform is vital in today's uncertain market," said Scott Sherwood, Novi's CEO.

About Novi Labs Novi Labs, Inc. ("Novi") is the leading developer of artificial intelligence driven business applications that help the oil & gas industry optimize the economic value of drilling programs and acquisition & divestiture decisions. Leveraging cutting-edge data science, Novi delivers intuitive analytics that simplify complex decisions with actionable data and insights needed optimize capital allocation. Novi was founded in 2014 and is headquartered in Austin, TX. For more information, please visit http://www.novilabs.com.

SOURCE Novi Labs

See the rest here:
Novi Releases v2.0 of Prediction Engine, Adding Critical Economics to Its Machine Learning Outputs - Benzinga

COVID-19 will change how the majority of us live and work, at least in the short term. Its also creating a challenge for tech companies such as Facebook, Twitter and Google that ordinarily rely on lots and lots of human labor to moderate content. Are A.I. and machine learning advanced enough to help these firms handle the disruption?

First, its worth noting that, although Facebook has instituted a sweeping work-from-home policy in order to protect its workers (along with Googleand a rising number of other firms), it initially required its contractors who moderate content to continue to come into the office. That situation only changed after protests,according toThe Intercept.

Now, Facebook is paying those contractors while they sit at home, since the nature of their work (scanning peoples posts for content that violates Facebooks terms of service) is extremely privacy-sensitive. Heres Facebooks statement:

For both our full-time employees and contract workforce there is some work that cannot be done from home due to safety, privacy and legal reasons. We have taken precautions to protect our workers by cutting down the number of people in any given office, implementing recommended work from home globally, physically spreading people out at any given office and doing additional cleaning. Given the rapidly evolving public health concerns, we are taking additional steps to protect our teams and will be working with our partners over the course of this week to send all contract workers who perform content review home, until further notice. Well ensure that all workers are paid during this time.

Facebook, Twitter, Reddit, and other companies are in the same proverbial boat: Theres an increasing need to police their respective platforms, if only to eliminate fake news about COVID-19, but the workers who handle such tasks cant necessarily do so from home, especially on their personal laptops. The potential solution? Artificial intelligence (A.I.) and machine-learning algorithms meant to scan questionable content and make a decision about whether to eliminate it.

HeresGoogles statement on the matter, via its YouTube Creator Blog.

Our Community Guidelines enforcement today is based on a combination of people and technology: Machine learning helps detect potentially harmful content and then sends it to human reviewers for assessment. As a result of the new measures were taking, we will temporarily start relying more on technology to help with some of the work normally done by reviewers. This means automated systems will start removing some content without human review, so we can continue to act quickly to remove violative content and protect our ecosystem, while we have workplace protections in place.

To be fair, the tech industry has been heading in this direction for some time. Relying on armies of human beings to read through every piece of content on the web is expensive, time-consuming, and prone to error. But A.I. and machine learning are still nascent, despite the hype. Google itself, in the aforementioned blog posting, pointed out how its automated systems may flag the wrong videos. Facebook is also receiving criticism that its automated anti-spam system is whacking the wrong posts, including those thatoffer vital information on the spread of COVID-19.

If the COVID-19 crisis drags on, though, more companies will no doubt turn to automation as a potential solution to disruptions in their workflow and other processes. That will force a steep learning curve; again and again, the rollout of A.I. platforms has demonstrated that, while the potential of the technology is there, implementation is often a rough and expensive processjust look at Google Duplex.

Membership has its benefits. Sign up for a free Dice profile, add your resume, discover great career insights and set your tech career in motion. Register now

Nonetheless, an aggressive embrace of A.I. will also create more opportunities for those technologists who have mastered A.I. and machine-learning skills of any sort; these folks may find themselves tasked with figuring out how to automate core processes in order to keep businesses running.

Before the virus emerged, BurningGlass (which analyzes millions of job postings from across the U.S.), estimated that jobs that involve A.I. would grow 40.1 percent over the next decade. That percentage could rise even higher if the crisis fundamentally alters how people across the world live and work. (The median salary for these positions is $105,007; for those with a PhD, it drifts up to $112,300.)

If youre trapped at home and have some time to learn a little bit more about A.I., it could be worth your time to explore online learning resources. For instance, theres aGooglecrash coursein machine learning. Hacker Noonalso offers an interesting breakdown ofmachine learningandartificial intelligence.Then theres Bloombergs Foundations of Machine Learning,a free online coursethat teaches advanced concepts such as optimization and kernel methods.

Read more from the original source:
Will COVID-19 Create a Big Moment for AI and Machine Learning? - Dice Insights

Introduction

In 2019, over 40 million Americans wagered money on March Madness brackets, according to the American Gaming Association. Most of this money was bet in bracket pools, which consist of a group of people each entering their predictions of the NCAA tournament games along with a buy-in. The bracket that comes closest to being right wins. If you also consider the bracket pools where only pride is at stake, the number of participants is much greater. Despite all this attention, most do not give themselves the best chance to win because they are focused on the wrong question.

The Right Question

Mistake #3 in Dr. John Elders Top 10 Data Science Mistakes is to ask the wrong question. A cornerstone of any successful analytics project starts with having the right project goal; that is, to aim at the right target. If youre like most people, when you fill out your bracket, you ask yourself, What do I think is most likely to happen? This is the wrong question to ask if you are competing in a pool because the objective is to win money, NOT to make the most correct bracket. The correct question to ask is: What bracket gives me the best chance to win $? (This requires studying the payout formula. I used ESPN standard scoring (320 possible points per round) with all pool money given to the winner. (10 points are awarded for each correct win in the round of 64, 20 in the round of 32, and so forth, doubling until 320 are awarded for a correct championship call.))

While these questions seem similar, the brackets they produce will be significantly different.

If you ignore your opponents and pick the teams with the best chance to win games you will reduce your chance of winning money. Even the strongest team is unlikely to win it all, and even if they do, plenty of your opponents likely picked them as well. The best way to optimize your chances of making money is to choose a champion team with a good chance to win who is unpopular with your opponents.

Knowing how other people in your pool are filling out their brackets is crucial, because it helps you identify teams that are less likely to be picked. One way to see how others are filling out their brackets is via ESPNs Who Picked Whom page (Figure 1). It summarizes how often each team is picked to advance in each round across all ESPN brackets and is a great first step towards identifying overlooked teams.

Figure 1. ESPNs Who Picked Whom Tournament Challenge page

For a team to be overlooked, their perceived chance to win must be lower than their actual chance to win. The Who Picked Whom page provides an estimate of perceived chance to win, but to find undervalued teams we also need estimates for actual chance to win. This can range from a complex prediction model to your own gut feeling. Two sources I trust are 538s March Madness predictions and Vegas future betting odds. 538s predictions are based on a combination of computer rankings and has predicted performance well in past tournaments. There is also reason to pay attention to Vegas odds, because if they were too far off, the sportsbooks would lose money.

However, both sources have their flaws. 538 is based on computer ratings, so while they avoid human bias, they miss out on expert intuition. Most Vegas sportsbooks likely use both computer ratings and expert intuition to create their betting odds, but they are strongly motivated to have equal betting on all sides, so they are significantly affected by human perception. For example, if everyone was betting on Duke to win the NCAA tournament, they would increase Dukes betting odds so that more people would bet on other teams to avoid large losses. When calculating win probabilities for this article, I chose to average 538 and Vegas predictions to obtain a balance I was comfortable with.

Lets look at last year. Figure 2 compares a teams perceived chance to win (based on ESPNs Who Picked Whom) to their actual chance to win (based on 538-Vegas averaged predictions) for the leading 2019 NCAA Tournament teams. (Probabilities for all 64 teams in the tournament appear in Table 6 in the Appendix.)

Figure 2. Actual versus perceived chance to win March Madness for 8 top teams

As shown in Figure 2, participants over-picked Duke and North Carolina as champions and under-picked Gonzaga and Virginia. Many factors contributed to these selections; for example, most predictive models, avid sports fans, and bettors agreed that Duke was the best team last year. If you were the picking the bracket most likely to occur, then selecting Duke as champion was the natural pick. But ignoring selections made by others in your pool wont help you win your pool.

While this graph is interesting, how can we turn it into concrete takeaways? Gonzaga and Virginia look like good picks, but what about the rest of the teams hidden in that bottom left corner? Does it ever make sense to pick teams like Texas Tech, who had a 2.6% chance to win it all, and only 0.9% of brackets picking them? How much does picking an overvalued favorite like Duke hurt your chances of winning your pool?

To answer these questions, I simulated many bracket pools and found that the teams in Gonzagas and Virginias spots are usually the best picksthe most undervalued of the top four to five favorites. However, as the size of your bracket pool increases, overlooked lower seeds like third-seeded Texas Tech or fourth-seeded Virginia Tech become more attractive. The logic for this is simple: the chance that one of these teams wins it all is small, but if they do, then you probably win your pool regardless of the number of participants, because its likely no one else picked them.

Simulations Methodology

To simulate bracket pools, I first had to simulate brackets. I used an average of the Vegas and 538 predictions to run many simulations of the actual events of March Madness. As discussed above, this method isnt perfect but its a good approximation. Next, I used the Who Picked Whom page to simulate many human-created brackets. For each human bracket, I calculated the chance it would win a pool of size by first finding its percentile ranking among all human brackets assuming one of the 538-Vegas simulated brackets were the real events. This percentile is basically the chance it is better than a random bracket. I raised the percentile to the power, and then repeated for all simulated 538-Vegas brackets, averaging the results to get a single win probability per bracket.

For example, lets say for one 538-Vegas simulation, my bracket is in the 90th percentile of all human brackets, and there are nine other people in my pool. The chance I win the pool would be. If we assumed a different simulation, then my bracket might only be in the 20th percentile, which would make my win probability . By averaging these probabilities for all 538-Vegas simulations we can calculate an estimate of a brackets win probability in a pool of size , assuming we trust our input sources.

Results

I used this methodology to simulate bracket pools with 10, 20, 50, 100, and 1000 participants. The detailed results of the simulations are shown in Tables 1-6 in the Appendix. Virginia and Gonzaga were the best champion picks when the pool had 50 or fewer participants. Yet, interestingly, Texas Tech and Purdue (3-seeds) and Virginia Tech (4-seed) were as good or better champion picks when the pool had 100 or more participants.

General takeaways from the simulations:

Additional Thoughts

We have assumed that your local pool makes their selections just like the rest of America, which probably isnt true. If you live close to a team thats in the tournament, then that team will likely be over-picked. For example, I live in Charlottesville (home of the University of Virginia), and Virginia has been picked as the champion in roughly 40% of brackets in my pools over the past couple of years. If you live close to a team with a high seed, one strategy is to start with ESPNs Who Picked Whom odds, and then boost the odds of the popular local team and correspondingly drop the odds for all other teams. Another strategy Ive used is to ask people in my pool who they are picking. It is mutually beneficial, since Id be less likely to pick whoever they are picking.

As a parting thought, I want to describe a scenario from the 2019 NCAA tournament some of you may be familiar with. Auburn, a five seed, was winning by two points in the waning moments of the game, when they inexplicably fouled the other team in the act of shooting a three-point shot with one second to go. The opposing player, a 78% free throw shooter, stepped to the line and missed two out of three shots, allowing Auburn to advance. This isnt an alternate reality; this is how Auburn won their first-round game against 12-seeded New Mexico State. They proceeded to beat powerhouses Kansas, North Carolina, and Kentucky on their way to the Final Four, where they faced the exact same situation against Virginia. Virginias Kyle Guy made all his three free throws, and Virginia went on to win the championship.

I add this to highlight an important qualifier of this analysisits impossible to accurately predict March Madness. Were the people who picked Auburn to go to the Final Four geniuses? Of course not. Had Terrell Brown of New Mexico State made his free throws, they would have looked silly. There is no perfect model that can predict the future, and those who do well in the pools are not basketball gurus, they are just lucky. Implementing the strategies talked about here wont guarantee a victory; they just reduce the amount of luck you need to win. And even with the best modelsyoull still need a lot of luck. It is March Madness, after all.

Appendix: Detailed Analyses by Bracket Sizes

At baseline (randomly), a bracket in a ten-person pool has a 10% chance to win. Table 1 shows how that chance changes based on the round selected for a given team to lose. For example, brackets that had Virginia losing in the Round of 64 won a ten-person pool 4.2% of the time, while brackets that picked them to win it all won 15.1% of the time. As a reminder, these simulations were done with only pre-tournament informationthey had no data indicating that Virginia was the eventual champion, of course.

Table 1 Probability that a bracket wins a ten-person bracket pool given that it had a given team (row) making it to a given round (column) and no further

In ten-person pools, the best performing brackets were those that picked Virginia or Gonzaga as the champion, winning 15% of the time. Notably, early round picks did not have a big influence on the chance of winning the pool, the exception being brackets that had a one or two seed losing in the first round. Brackets that had a three seed or lower as champion performed very poorly, but having lower seeds making the Final Four did not have a significant impact on chance of winning.

Table 2 shows the same information for bracket pools with 20 people. The baseline chance is now 5%, and again the best performing brackets are those that picked Virginia or Gonzaga to win. Similarly, picks in the first few rounds do not have much influence. Michigan State has now risen to the third best Champion pick, and interestingly Purdue is the third best runner-up pick.

Table 2 Probability that a bracket wins a 20-person bracket pool given that it had a given team (row) making it to a given round (column) and no further

When the bracket pool size increases to 50, as shown in Table 3, picking the overvalued favorites (Duke and North Carolina) as champions significantly lowers your baseline chances (2%). The slightly undervalued two and three seeds now raise your baseline chances when selected as champions, but Virginia and Gonzaga remain the best picks.

Table 3 Probability that a bracket wins a 50-person bracket pool given that it had a given team (row) making it to a given round (column) and no further

With the bracket pool size at 100 (Table 4), Virginia and Gonzaga are joined by undervalued three-seeds Texas Tech and Purdue. Picking any of these four raises your baseline chances from 1% to close to 2%. Picking Duke or North Carolina again hurts your chances.

Table 4 Probability that a bracket wins a 100-person bracket pool given that it had a given team (row) making it to a given round (column) and no further

When the bracket pool grows to 1000 people (Table 5), there is a complete changing of the guard. Virginia Tech is now the optimal champion pick, raising your baseline chance of winning your pool from 0.1% to 0.4%, followed by the three-seeds and sixth-seeded Iowa State are the best champion picks.

Table 5 Probability that a bracket wins a 1000-person bracket pool given that it had a given team (row) making it to a given round (column) and no further

For Reference, Table 6 shows the actual chance to win versus the chance of being picked to win for all teams seeded seventh or better. These chances are derived from the ESPN Who Picked Whom page and the 538-Vegas predictions. The data for the top eight teams in Table 6 is plotted in Figure 2. Notably, Duke and North Carolina are overvalued, while the rest are all at least slightly undervalued.

The teams in bold in Table 6 are examples of teams that are good champion picks in larger pools. They all have a high ratio of actual chance to win to chance of being picked to win, but a low overall actual chance to win.

Table 6 Actual odds to win Championship vs Chance Team is Picked to Win Championship.

Undervalued teams in green; over-valued in red.

About the Author

Robert Robison is an experienced engineer and data analyst who loves to challenge assumptions and think outside the box. He enjoys learning new skills and techniques to reveal value in data. Robert earned a BS in Aerospace Engineering from the University of Virginia, and is completing an MS in Analytics through Georgia Tech.

In his free time, Robert enjoys playing volleyball and basketball, watching basketball and football, reading, hiking, and doing anything with his wife, Lauren.

Read the original:
How to Pick a Winning March Madness Bracket - Machine Learning Times - machine learning & data science news - The Predictive Analytics Times

Researchers have been working to make mind-controlled prosthetics a reality for at least a decade. In theory, an artificial hand that amputees could control with their mind could restore their ability to carry out all sorts of daily tasks, and dramatically improve their standard of living.

However, until now scientists have faced a major barrier: they havent been able to access nerve signals that are strong or stable enough to send to the bionic limb. Although its possible to get this sort of signal using a brain-machine interface, the procedure to implant one is invasive and costly. And the nerve signals carried by the peripheral nerves that fan out from the brain and spinal cord are too small.

A new implant gets around this problem by using machine learning to amplify these signals. A study, published in Science Translational Medicine today, found that it worked for four amputees for almost a year. It gave them fine control of their prosthetic hands and let them pick up miniature play bricks, grasp items like soda cans, and play Rock, Paper, Scissors.

Sign up for The Algorithm artificial intelligence, demystified

Its the first time researchers have recorded millivolt signals from a nervefar stronger than any previous study.

The strength of this signal allowed the researchers to train algorithms to translate them into movements. The first time we switched it on, it worked immediately, says Paul Cederna, a biomechanics professor at the University of Michigan, who co-led the study. There was no gap between thought and movement.

The procedure for the implant requires one of the amputees peripheral nerves to be cut and stitched up to the muscle. The site heals, developing nerves and blood vessels over three months. Electrodes are then implanted into these sites, allowing a nerve signal to be recorded and passed on to a prosthetic hand in real time. The signals are turned into movements using machine-learning algorithms (the same types that are used for brain-machine interfaces).

Amputees wearing the prosthetic hand were able to control each individual finger and swivel their thumbs, regardless of how recently they had lost their limb. Their nerve signals were recorded for a few minutes to calibrate the algorithms to their individual signals, but after that each implant worked straight away, without any need to recalibrate during the 300 days of testing, according to study co-leader Cynthia Chestek, an associate professor in biomedical engineering at the University of Michigan.

Its just a proof-of-concept study, so it requires further testing to validate the results. The researchers are recruiting amputees for an ongoing clinical trial, funded by DARPA and the National Institutes of Health.

Read the original post:
An implant uses machine learning to give amputees control over prosthetic hands - MIT Technology Review

GREENWOOD VILLAGE, Colo., Feb. 17, 2020 /PRNewswire-PRWeb/ --Driven by a passion for learning and all things data science, Brian Burch has cultivated an exemplary career in building solutions which solve business problems across multiple industries including cybersecurity, financial services, retail, telecommunications, and aerospace. In addition to having a strong technical background across a broad range of vertical markets, Brian brings deep expertise in the areas of Artificial Intelligence and Machine Learning (AI/ML), Software Engineering, and Product Management.

"We are excited about Brian Burch joining the zvelo leadership team," explains zvelo CEO, Jeff Finn. "zvelo is quickly gaining momentum with tremendous growth opportunities built upon the zveloAI platform. Brian brings an impressive background in AI/ML and data science to further zvelo's leadership for URL classification, objectionable and malicious detection and his passion aligns perfectly with zvelo's mission to improve internet safety and security."

From large organizations like CenturyLink and Regions Bank to successful startups like StorePerform Technologies and Cognilytics, Brian has a proven history of leveraging his vast experience in key leadership roles to advance business goals through a fully-immersed, hands-on approach.

"I'm especially excited about combining zvelo's strong web categorization technologies with the latest advances in AI/ML to identify malicious websites, phishing URLs, and malware distribution infrastructure, and play a key role in supporting the mission to make the internet safer for everyone," stated Burch.

About zvelo, Inc. zvelo is a leading provider of web content classification and detection of objectionable, malicious and threat detection services with a mission of making the Internet safer and more secure. zvelo combines advanced artificial intelligence-based contextual categorization with sophisticated malicious and phishing detection capabilities that customers integrate into network and endpoint security, URL and DNS filtering, brand safety, contextual targeting, and other applications where data quality, accuracy, and detection rates are critical.

Learn more at: https://www.zvelo.com

Corporate Information: zvelo, Inc. 8350 East Crescent Parkway, Suite 450 Greenwood Village, CO 80111 Phone: (720) 897-8113 zvelo.com or pr@zvelo.com

SOURCE zvelo

More here:
Brian Burch Joins zvelo as Head of Artificial Intelligence and Machine Learning to Drive New Growth Initiatives - Benzinga

Apple thinks it can improve location accuracy by applying machine learning to Kalman estimation filters, a just-published patent application reveals. Kalman filters are popularly used in GPS and robotic motion tracking applications. And, now Apple wants to use machine learning along with Kalman filters to bring the accuracy of positioning down to centimetre-level.

According to the patent application, Apple proposes:

The device, say an iPhone would generate a machine learning model, for example, by comparing GNSS position estimates (or estimated measurement errors) with corresponding reference position estimates (where the reference positions correspond to ground truth data).

In one or more implementations, the ground truth data may be better (e.g., significantly better) than what a mobile device alone can perform in most non-aided mode(s) of operation. For example, a mobile phone in a car may be significantly better aided than a pedestrian device, because the motion model for a vehicle is more constrained, and has aiding data in the form of maps and sensor inputs.

Tall buildings and tree cover can confuse the positioning systems to accurately locate the user. So, Apple wants to generate machine learning models on the device that would predict the users location based on its training as well as a reference position.

Today even Apples rivals praise Apple for it has done to the electronics industry. In an in-depth CNBC interview with Huaweis founder and CEO Ren Zhengfei earlier this year he spoke about how Apple has revolutionised the era of the Internet. In his ascent, however, Apple has put many traditional companies to dust.

According to a 2018 CNBCs report, there has been a dramatic decline in worldwide shipments of cameras. The chart above from Statista illustrates this fall, which also coincides with the peaking of Apple iPhone and its ever-improving camera.

So, the companies those who outsource their GPS improving services will be watching the new ML-based GPS patent closely or even might be rushing to build something of their own. However, this might not be the case in this modern era of mega collaborations.

Last month we saw one of the biggest corporate crossovers of the 21st century, when the tech giants, Amazon, Apple and Google, along with others announced their plans to develop compatible smart home products together.

Gone are the days where companies build something up from scratch (with the exception of Tesla). If your rival company is good at something you are not, then you either buy a startup that works solely on that technology or join hands with the rival. So, Apples patent to improve GPS in the upcoming 5G era might receive a warm welcome.

Of course, there always will be a debate about whether one should patent widely used technology, which can hand over infinite leverage to a single entity.

That said, the last two years has since increased attention in seeking patents over ML-based techniques. Last year, it was Google, which has been in the news for patenting machine learning techniques such as batch normalisation. Companies like Google and Apple have been leading the AI race for quite some time. It can also be possible if it is a routine to apply for a patent for their innovations and this new-found obsession over ML patent news is due to the rising popularity of AI globally.

At the end of the day, it comes down to whether you should risk years worth of intellectual property to a potential patent troll or safeguard it through patenting and then democratise the technology to the masses. It has been the latter, for many years and we have to wait and watch if machine learning-based patents find an exception as we go forward.

comments

Read more from the original source:
Apple Patents ML Based Navigation System For Its Maps, But Why? - Analytics India Magazine

Social media can send up an early warning sign of illness, and data analysis can predict how it will spread.

The coronavirus illness (nCoV) is now an international public health emergency, bigger than the SARS outbreak of 2003. Unlike SARS, this time around scientists have better genome sequencing, machine learning, and predictive analysis tools to understand and monitor the outbreak.

During the SARS outbreak, it took five months for scientists to sequence the virus's genome. However, the first 2019-nCoV case was reported in December, and scientists had the genome sequenced by January 10, only a month later.

Researchers have been using mapping tools to track the spread of disease for several years. Ten European countries started Influenza Net in 2003 to track flu symptoms as reported by individuals, and the American version, Flu Near You, started a similar service in 2011.

Lauren Gardner, a civil engineering professor at Johns Hopkins and the co-director of the Center for Systems Science and Engineering, led the effort to launch a real-time map of the spread of the 2019-nCoV. The site displays statistics about deaths and confirmed cases of coronavirus on a worldwide map.

Este Geraghty, MD, MS, MPH, GISP, and chief medical officer and health solutions director at Esri, said that since the SARS outbreak in 2003 there has been a revolution in applied geography through web-based tools.

"Now as we deploy these tools to protect human lives, we can ingest real-time data and display results in interactive dashboards like the coronavirus dashboard built by Johns Hopkins University using ArcGIS," she said.

SEE:The top 10 languages for machine learning hosted on GitHub (free PDF)

With this outbreak, scientists have another source of data that did not exist in 2003: Twitter and Facebook. In 2014, Chicago's Department of Innovation and Technology built an algorithm that used social media mining and illness prediction technologies to target restaurants inspections. It worked: The algorithm found violations about 7.5 days before the normal inspection routine did.

Theresa Do, MPH, leader of the Federal Healthcare Advisory and Solutions team at SAS, said that social media can be used as an early indicator that something is going on.

"When you're thinking on a world stage, a lot of times they don't have a lot of these technological advances, but what they do have is cell phones, so they may be tweeting out 'My whole village is sick, something's going on here,' she said.

Do said an analysis of social media posts can be combined with other data sources to predict who is most likely to develop illnesses like the coronavirus illness.

"You can use social media as a source but then validate it against other data sources," she said. "It's not always generalizable (is generalizable a word?), but it can be a sentinel source."

Do said predictive analytics has made significant advances since 2003, including refining the ability to combine multiple data sources. For example, algorithms can look at names on plane tickets and compare that information with data from other sources to predict who has been traveling to certain areas.

"Algorithms can allow you to say 'with some likelihood' it's likely to be the same person," she said.

The current challenge is identifying gaps in the data. She said that researchers have to balance between the need for real-time data and privacy concerns.

"If you think about the different smartwatches that people wear, you can tell if people are active or not and use that as part of your model, but people aren't always willing to share that because then you can track where someone is at all times," she said.

Do said that the coronavirus outbreak resembles the SARS outbreak, but that governments are sharing data more openly this time.

"We may be getting a lot more positives than they're revealing and that plays a role in how we build the models," she said. "A country doesn't want to be looked at as having the most cases but that is how you save lives."

Get expert tips on mastering the fundamentals of big data analytics, and keep up with the latest developments in artificial intelligence. Delivered Mondays

This map from Johns Hopkins shows reported cases of 2019-nCoV as of January 30, 2020 at 9:30 pm. The yellow line in the graph is cases outside of China while the orange line shows reported cases inside the country.

Image: 2019-nCoV Global Cases by Johns Hopkins Center for Systems Science and Engineering

Follow this link:
Combating the coronavirus with Twitter, data mining, and machine learning - TechRepublic

(credit: Pablo Lagarto / Adobe Stock)

On 1 February 2020, YouTuber Simon Weekert posted a video on YouTube claiming to have redirected traffic by faking traffic jams on Google Maps. The video shows Weekert walking slowly along traffic-free streets in Berlin, pulling a pile of second-hand mobile phones in a cart behind him and Google Maps generating traffic jam alerts because the phones had their location services turned on.

Weekerts performance act demonstrates the fragility and vulnerability of our systems and their difficulty in interpreting outliers, and highlights a kind of decisional blindness when we think of data as objective, unambiguous and interpretation free, as he put it. There are many other examples of decisional blindness relating to drivers following Google Maps and falling off cliffs or driving into rivers.

Google has the resources, expertise and technology to rapidly learn from this experience and make changes to avoid similar situations. But the same vulnerability to hacking or outliers applies to the use of machine learning in childrens social care (CSC) and this raises the question of whether the sector has the means to identity and rectify issues in a timely manner and without adverse effects for service users.

Have you ever had the experience of asking the wrong question in Google search and getting the right answer? Thats because of contextual computing that makes use of AI and machine learning.

At its heart, machine learning is the application of statistical techniques to identify patterns and enable computers to use data to progressively learn and improve their performance.

From Google search and Alexa to online shopping, and from games and health apps to WhatsApp and online dating, most online interactions are mediated by AI and machine learning. Like electricity, AI and machine learning will power every software and digital device and will transform and mediate every aspect of human experience mostly without end users giving them a thought.

But there are particular concerns about their applications in CSC and, therefore, a corresponding need for national standards for machine learning in social care and for greater transparency and scrutiny around the purpose, design, development, use, operation and ethics of machine learning in CSC. This was set out in What Works for Childrens Social Cares ethics review into machine learning, published at the end of January.

The quality of machine learning systems predictive analysis is dependent on the quality, completeness and representativeness of the dataset they draw on. But peoples lives are complex, and often case notes do not capture this complexity and instead are complemented by practitioners intuition and practice wisdom. Such data lacks the quality and structure needed for machine learning applications, making high levels of accuracy harder to achieve.

Inaccuracy in identifying children and families can result in either false positives that infringe on peoples rights and privacy, cause stress and waste time and resources, or false negatives that miss children and families in need of support and protection.

Advocates of machine learning often point out that systems only provide assistance and recommendations, and that it remains the professionals who make actual decisions. Yet decisional blindness can undermine critical thinking, and false positives and negatives can result in poor practice and stigmatisation, and can further exclusion, harm and inequality.

Its true that AI and machine learning can be used in empowering ways to support services or to challenge discrimination and bias. The use of Amazons Alexa to support service users in adult social care is, while not completely free of concerns, one example of positive application of AI in practice.

Another is Essex councils use of machine learning to produce anonymised aggregate data at community level of children who may not be ready for school by their fifth birthday. This data is then shared with parents and services who are part of the project to inform their funding allocation or changes to practice as need be. This is a case of predictive analytics being used in a way that is supportive of children and empowering for parents and professionals.

The Principal Children and Families Social Worker (PCFSW) Network is conducting a survey of practitioners to understand their current use of technology and challenges and the skills, capabilities and support that they need.

It only takes 10 minutes to complete the survey on digital professionalism and online safeguarding. Your responses will inform best practice and better support for social workers and social care practitioners to help ensure practitioners lead the changes in technology rather than technology driving practice and shaping practitioners professional identity.

But its more difficult to make such an assessment in relation to applications that use hundreds of thousands of peoples data, without their consent, to predict child abuse. While there are obvious practical challenges around seeking the permission of huge numbers of people, failing to do so shifts the boundaries of individual rights and privacy vis--vis surveillance and the power of public authorities. Unfortunately though, ethical concerns do not always influence the direction or speed of change.

Another controversial recent application of technology is the use of live facial recognition cameras in London. An independent report by Essex Universitylast year suggested concerns with inaccuracies in use of live facial recognition, while the Met Polices senior technologist, Johanna Morley said millions of pounds would need to be invested in purging police suspect lists and aligning front- and back-office systems to ensure the legality of facial recognition cameras. Despite these concerns, the Met will begin using facial recognition cameras in London streets, with the aim of tackling serious crime, including child sexual exploitation.

Research published in November 2015, meanwhile, showed that a flock of trained pigeons can spot cancer in images of biopsied tissue with 99% accuracy; that is comparable to what would be expected of a pathologist. At the time, one of the co-authors of the report suggested that the birds might be able to assess the quality of new imaging techniques or methods of processing and displaying images without forcing humans to spend hours or days doing detailed comparisons.

Although there are obvious cost efficiencies in recruiting pigeons instead of humans, I am sure most of us will not be too comfortable having a flock of pigeons as our pathologist or radiologist.

Many people would also argue more broadly that fiscal policy should not undermine peoples health and wellbeing. Yet the past decade of austerity, with 16bn in cuts in core government funding for local authorities by this year and a continued emphasis on doing more with less, has led to resource-led practices that are far from the aspirations of Children Act 1989 and of every child having the opportunity to achieve their potential.

Technology is never neutral and there are winners and losers in every change. Given the profound implications of AI and machine learning for CSC, it is essential such systems are accompanied by appropriate safeguards and processes that prevent and mitigate false positives and negatives and their adverse impact and repercussions. But in an environment of severe cost constraints, positive aspirations might not be matched with adequate funding to ensure effective prevention and adequate support for those negatively impacted by such technologies.

In spite of the recent ethics reviews laudable aspirations, there is also the real risk that many of the applications of machine learning pursued to date in CSC may cement current practice challenges by hard-coding austerity and current thresholds into systems and the future of services.

The US constitution was written and ratified by middle-aged white men and it took over 130 years for women to gain the right of suffrage and 176 years to recognise and outlaw discrimination based on race, sex, religion and national origin. Learning from history would suggest we must be cautious about reflecting childrens social cares operating context into systems, all designed, developed and implemented by experts and programmers who may not represent the diversity of the people who will be most affected by such systems.

Dr Peter Buzzi (@MHChat) is the director of Research and Management Consultancy Centre and the Safeguarding Research Institute. He is also the national research lead for the Principal Children and Families Social Worker (PCFSW) Networks online safeguarding research and practice development project.

View post:
'Technology is never neutral': why we should remain wary of machine learning in children's social care - Communitycare.co.uk

Human resources has been slower to come to the table with machine learning and artificial intelligence than other fieldsmarketing, communications, even health care. But the value of machine learning in human resourcescan now be measured, thanks to advances in algorithms that can predict employee attrition, for example, or deep learning neural networks thatare edging toward more transparent reasoning in showing why a particular result or conclusion was made.

The value beyond numbers for CEOs and managersis the power inunderstanding whats actually happening within acompany i.e. withtheir people. AsGlintsJustin Black articulated in awebinar for the Human Capital Institute(HCI), executives and leaders need information that helps them point people in the right direction; informationsales data, KPIs, etc.change over time, and machine learning can react faster than people in helping draw out the insights and inferences that might otherwise take reams of manpower or not be uncovered at all.

Though not an exhaustive list, belowis an outline of solid examples of machine learning and artificial intelligence applications at work in human resources today, along with developing and near-future applications.

Applicant Tracking & Assessment

Applicant tracking and assessment has topped the list in early machine learning applications, especially for companies and roles that receive high volumes of applicants.Glintis not an AI company, but they use AI tools to help companies save money and provide a better work experience. Machine learning tools help HR and management personnel hirenew team members bytrackinga candidates journey throughout the interview process and helping speed up the processof getting streamlined feedback to applicants.

Peopliseis another solutionfor helpingcompanies calculate fit score for new talent, combining tools like digital screening and online interview results to help hiring managers arrive at decisions.

While competition for the best people has driven many HR departments to use algorithmic-based assessments, aCEBarticle on using machine learning to eliminate bias cautions that human oversight isstill of paramount importance. Its not enough to act directly on data insights, but to use this information in tandem with driving question such as: 1) how I can link applicant traits to business outcomes; 2) which outcomes should be our focus when hiring; and 3) can predictions (hiring and otherwise) be made in an unbiased way.

Attracting Talent

Attracting talent beforehiring has also seen an upswingin machine-learning based applications in the past few years. Black, who is Glints senior director of Organizational Development, named LinkedIn as an example of a company using one of the most common versions of basic machine learningrecommendingjobs. Other job-finding sites, including Indeed, Glassdoor, and Seek use similar algorithms to build interaction mapsbased on users data from previous searches, connections, posts, and clicks.

PhenomPeople is one example of a suite of machine learning-based toolsthat helps leadpotential talent to a companys career site through multiple social media and job search channels. Black notes that this is really just one step past a keyword search, albeit a big step computationally, as theres a lot more to do.

Attrition Detection

Understanding people and why they decide to stay at or leave a job is arguably one of the most important questions for HR to answer. Identifying attrition risk calls for advancedpattern recognition in surveying an array of variables.

In the earlier mentioned HCI webinar, Black describes a hypotheticalsituation of identifying specific risk factors based on scores to an employee survey. If a human were to try and detect attrition risk among female engineers in Palo Alto with less than 2 years of tenure, the variance analyses to reach that conclusion are innumerable, like finding a needle in haystack, but machine learning allows us to connect these dots in seconds, freeing HR representatives to spend time supporting teams instead of analyzing data.

Glints employee engagement platform

Advances in NLP have included the ability to process large amounts of unstructured data, and algorithms can also do things like identify emotional activity in comments and tease out prescriptive comments, or actionable suggestions. Black describesprototypicality algorithms that can pull out individual comments thatrepresent the sum of what everyones saying, allowing companies to get a broadly inclusive but digestible pulse on company processes and specific issues.

JPMorgan is apparently one of several financial institutions that hasalso put into place algorithms that can survey employee behavior and identify rogue employees before any criminal activity takes place, an obviously more insidious form of attrition with dire consequenceswatch the interview with Bloomberg Reporter Hugh Son as hediscussesthese new safeguards with Bloomberg Technology.

Individual Skills Management/Performance Development

Machine learning is showing its potential inboosting individual skill management and development. While there is definitely room for growth in this arena, platforms that cangive calibrated guidance without human coaches save time and provide the opportunity for more people togrow in their careers and stay engaged.Workday is just one example of a company building personalized training recommendations for employees based on a companys needs, market trends, and employee specifics.

Black elaborates that these types of performance development assessments are useful when actually read, which is why this type of machine-based feedback has been successful for individuals. But this becomes more difficult at the level of the organization, where its almost impossible to make sense of enormous amounts of varying data; this is an area wheremachine learning is evolving, with an increased focus on the overall performance of the corporate lattice.

Enterprise Management

As alluded to in the last example, enterprise management and engagement based on machine learning insights is already here in early forms but has yet to be taken to scale. KPMG promotes its customized Intelligent Enterprise Approach, leveraging predictive analytics and big data management to help companies make business decisions that optimize key KPIs and other metrics.re:Work, which provides best workplace practices and ideas from Google and other leading organizations (including KPMG), is an excellent resource for staying up-to-date on new tools and case studies in this space.

Googles People Analytics department has been a pioneer in building performance-management engines at the enterprise level. From an early stage, the team (led by Prasad Setty) posed existing questionsfor example, whats the ideal size for a given team or departmentbut focused on finding new ways to use data in order to help answer these questions. In turn, People Analytics has helped pave the way for solving fundamental business problems related to the employee life cycle, with afocus on improving Googlers'production and overall wellness. Asoutline by Chris Derose for The Atlantic, over the last half of a decade, the team has produced insights that have led to improvements in company-wide actions, such as:

Post-Hire Outcome Algorithms

CEB notes that theideal hiring algorithm would predicta post-hire outcome (for example, reducing time taking customer service calls while keeping customer satisfaction high) rather than just matching job requirements with items on an employees resume or pre-hire assessment results.

The article goes on to note that its sometimes the counterintuitive aspectsthat predict job performance, informationthat a machine is better at findingthrough analysis than human inference. For example, CEB describes a model created for a call center representative role that linked call center experience to resultingpoor performance. While a link to the source or actual model would be helpful, the idea is interesting and reflects machine learnings strengths in invisiblepattern recognition

Internal Management

WhenTalent AnalyticsChief Scientist Pasha Roberts discussed the role of predictive analytics in human resource management with Emerj in 2016, he brought up the internal movement of employees within a company as an issue unique to HR and analytics. You can use agent-based modeling to simulate and look at how people can move within a companyand be better able to hire a person at the entry-level that will be likely to move through corporate ladder, said Roberts. While there are early systems in place, more data over time should lead to a more robust and scalable model for internal management over the nextfive years.

Increased Behavior Tracking and Data-Based Decision Making

Ben Waber, president and CEO ofHumanyzeand also a past guest on Emerj, talked about the increasing use of IoT wearable data in the workplace. These types of gadgets are more common at the enterprise levelbluetooth headphones and smart ID badges, for exampleand companies are continuing toadd sensor technology to the workplace in order to collect data. This is an area that Waber researched while serving as a visiting scientist at the MIT Media Lab, using data collected from smart badges to look at things like employee dialogue, interaction, networks within a company, where people spent their time, etc. It would seem that privacy might be a concern, but technologies like smart badges are starting to proliferate quickly (with vendors like Atmel, in the below video, introducing new and updated apps for Android phones). This type of data, says Waber, allows us to pose and answer crucial business-driving questions that we couldnt ask before, such as how much does my sales team talk to my engineering team?

Things to Keep in Mind:Machine Learning in Human Resources

Google People Analytics Lead, Ian OKeefe, told a story at the People Analytics & Future of Work conference inJanuary 2016 about his teams efforts to quantify things like efficiency, effectiveness and employee experience by looking at hiring decisions, teamclimate, and personal development. In the end, his team found that people armed with better data make better decisions than algorithms alone can do.

Well-designed AI applications, says Black, have three main cross functions: main expertise, data science expertise, and design/user experience expertise. At present, very few providers do all three of these well. The best solutions today and in the near futuredont replace humans, but emphasize scaling better decision making with the use of machines as a tool and collaborator.

Our survey of machine learning in human resources illuminates the development of a more people-centric approach, paving the way for more more valuable programs and less wasted time; reduced bias in programs; less administration and more individual development; and the ability to act proactively rather than reactively, moving seamlessly fromthe level of the individual to the organization and back again.

Image credit: Corporate IT

Read the original here:
Machine Learning in Human Resources Applications and ...

A couple of weeks after the Coronavirus outbreak and the disease has become a full-blown pandemic. According to official Chinese statistics, more than 130 people have died from the mysterious virus.

Contagious diseases may be diagnosed by men and women in face masks and lab coats, but warning signs of an epidemic can be detected by computer programmers sitting thousands of miles away. Around the tenth of January, news of a flu outbreak in Chinas Hubei province started making its way to mainstream media. It then spread to other parts of the country, and subsequently, overseas.

But the first to report of an impending biohazard was BlueDot, a Canadian firm that specializes in infectious disease surveillance. They predicted an impending outbreak of coronavirus on December 31 using an artificial intelligence-powered system that combs through animal and plant disease networks, news reports in vernacular websites, government documents, and other online sources to warn its clients against traveling to danger zones like Wuhan, much before foreign governments started issuing travel advisories.

They further used global airline ticketing data to correctly predict that the virus would spread to Seoul, Bangkok, Taipei, and Tokyo. Machine learning and natural language processing techniques were also employed to create models that process large amounts of data in real time. This includes airline ticketing data, news reports in 65 languages, animal and plant disease networks.

iStock

We know that governments may not be relied upon to provide information in a timely fashion. We can pick up news of possible outbreaks, little murmurs or forums or blogs of indications of some kind of unusual events going on, Kamran Khan, founder and CEO of BlueDot told a news magazine.

The death toll from the Coronavirus rose to 81 in China, with thousands of new cases registered each day. The government has extended the Lunar New Year holiday by three days to restrict the movement of people across the country, and thereby lower the chances of more people contracting the respiratory disease.

However, a lockdown of the affected area could be detrimental to public health, putting at risk the domestic population, even as medical supplies dwindle, causing much anger and resentment.

24 May, 2018

24 May, 2018

24 May, 2018

24 May, 2018

24 May, 2018

Read this article:
This tech firm used AI & machine learning to predict Coronavirus outbreak; warned people about danger zones - Economic Times

Artificial intelligence and machine learning will play an even bigger role in healthcare in 2020 than they did in 2019, helping medical professionals with everything from oncology screenings to note-taking.

On top of actual deployments, increased investment activity is also expected this year, and with deeper deployments of AI and ML technology, a broader base of test cases will be available to collect valuable best practices information.

As AI is implemented more widely in real-world clinical practice, there will be more academic reports on the clinical benefits that have arisen from the real-world use, said Pete Durlach, senior vice president for healthcare strategy and new business development at Nuance.

"With healthy clinical evidence, we'll see AI become more mainstream in various clinical settings, creating a positive feedback loop of more evidence-based research and use in the field," he explained. "Soon, it will be hard to imagine a doctor's visit, or a hospital stay that doesn't incorporate AI in numerous ways."

In addition, AI and ambient sensing technology will help re-humanize medicine by allowing doctors to focus less on paperwork and administrative functions, and more on patient care.

"As AI becomes more commonplace in the exam room, everything will be voice enabled, people will get used to talking to everything, and doctors will be able to spend 100% of their time focused on the patient, rather than entering data into machines," Durlach predicted. "We will see the exam room of the future where clinical documentation writes itself."

The adoption of AI for robotic process automation ("RPA") for common and high value administrative functions such as the revenue cycle, supply chain and patient scheduling also has the potential to rapidly increase as AI helps automate or partially automate components of these functions, driving significantly enhanced financial outcomes to provider organizations.

Durlach also noted the fear that AI will replace doctors and clinicians has dissipated, and the goal now is to figure out how to incorporate AI as another tool to help physicians make the best care decisions possible effectively augmenting the intelligence of the clinician.

"However, we will still need to protect against phenomenon like alert fatigue, which occurs when users who are faced with many low-level alerts, ignore alerts of all levels, thereby missing crucial ones that can affect the health and safety of patients," he cautioned.

In the next few years, he predicts the market will see a technology that finds a balance between being too obtrusive while supporting doctors to make the best decisions for their patients as the learn to trust the AI powered suggestions and recommendations.

"So many technologies claim they have an AI component, but often there's a blurred line in which the term AI is used in a broad sense, when the technology that's being described is actually basic analytics or machine learning," Kuldeep Singh Rajput, CEO and founder of Boston-based Biofourmis, told Healthcare IT News. "Health system leaders looking to make investments in AI should ask for real-world examples of how the technology is creating ROI for other organizations."

For example, he pointed to a study of Brigham & Women's Home Hospital program, recently published in Annals of Internal Medicine, which employed AI-driven continuous monitoring combined with advanced physiology analytics and related clinical care as a substitute for usual hospital care.

The study found that the program--which included an investment in AI-driven predictive analytics as a key component--reduced costs, decreased healthcare use, and lowered readmissions while increasing physical activity compared with usual hospital care.

"Those types of outcomes could be replicated by other healthcare organizations, which makes a strong clinical and financial case to invest in that type of AI," Rajput said.

Nathan Eddy is a healthcare and technology freelancer based in Berlin.Email the writer:nathaneddy@gmail.comTwitter:@dropdeaded209

Continued here:
AI and machine learning trends to look toward in 2020 - Healthcare IT News

For all the potential benefits of artificial intelligence and machine learning, one of the biggest and, increasingly most publicized challenges with the technology is the potential for algorithmic bias.

But an even more basic challenge for hospitals and health systems looking to deploy AI and ML can be the skepticism from frontline staff a hesitance to use predictive models that, even if they aren't inherently biased, are certainly hard to understand.

At Delaware-based Christiana Care Health System, the past few years have seen efforts to "simplify the model without sacrificing precision," says Dr. Terri Steinberg, its chief health information officer and VP of population health informatics.

"The simpler the model, the more human beings will accept it," said Steinberg, who will talk more about this notion in a March 12 presentation at HIMSS20.

When it comes to pop health programs, the data sets used to drive the analytics matter, she explains. Whether it's EHR data, social determinants of health, claims data or even wearables information, it's key to select the most relevant data sources, use machine learning to segment the population and then, crucially, present those findings to care managers in a way that's understandable and fits their workflow.

At HIMSS20, Steinberg, alongside Health Catalyst Chief Data Scientist Jason Jones, will show how Christiana Care has been working to streamline its machine learning processes, to ensure they're more approachable and thus more liable to be embraced by its care teams.

Dr. Terri Steinberg, Christiana Care Health System

They'll explain how to assign relative value to pop health data and discuss some of the challenges associated with integrating them; they'll show how ML can segment populations and spotlight strategies for using new data sources that will boost the value and utility of predictive models.

"We've been doing this since 2012," said Steinberg. And now we have significant time under our belts, so we wanted to come back to HIMSS and talk about what we were doing in terms of programming for care management and, more important, how we're segmenting our population with machine learning."

"There are a couple of patterns that we've seen repeated across engagements that are a little bit counter to how people typically go about building these models today, which is to sort of throw everything at them and hope for the best," said Jones, of Health Catalyst, Christiana Care's vendor partner.

At Christiana Care, he said, the goal instead has been to "help people understand as much as they would like about how the models are working, so that they will trust and actually use them.

"We've found repeatedly that we can build technically fantastic models that people just don't trust and won't use," he added. "In that case, we might as well not bother in the first place. So we're going to go through and show how it is that we can build models in such a way that they're technically excellent but also well-trusted by the people who are going to use them."

In years past, "when we built the model and put it in front of our care managers and said, 'Here you go, now customize your treatment plans based on the risk score,' what we discovered is that they basically ignored the score and did what they wanted," Steinberg explained.

But by simplifying a given model to the "smallest number of participants and data elements that can be," that enables the development of something "small enough for people to understand the list of components, so that they think that they know why the model has made a specific prediction," she said.

That has more value than many population health professionals realize.

"The goal is to simplify the model as much as you can, so human beings understand the components," said Steinberg.

"People like understanding why a particular individual falls into a risk category," she said. "And then they sometimes would even like to know what the feature is that has resulted in the risk. The take home message is that the more human beings understand what the machine is doing, the more likely they are to trust the machine. We want to personalize the black box."

Steinberg and Jones will talk more about making machine learning meaningful at a HIMSS20 session titled "Machine Learning and Data Selection for Population Health." It's scheduled for Thursday, March 12, from 10-11 a.m. in room W414A.

Continue reading here:
Christiana Care offers tips to 'personalize the black box' of machine learning - Healthcare IT News

Deep Learning Super Sampling (DLSS) is one of the marquee features for Nvidias RTX video cards, but its also one people tend to overlook or outright dismiss. The reason for that is because many people equate the technology to something like a sharpening filter that can sometimes reduce the jagged look of lower-resolution images. But DLSS uses a completely different method with much more potential for improving visual quality, and Nvidia is ready to prove that with DLSS 2.0.

Nvidia built the second-generation DLSS to address all of the concerns with the technology. It looks better, gives players much more control, and should support a lot more games. But at its core, DLSS 2.0 is still about using machine learning to intelligently upscale a game to a higher resolution. The idea is to give you a game that, for example, looks like it is running at 4K while actually rendering at 1080p or 1440p. This drastically improves performance. And, in certain games, it can even produce frames that contain more detail than native rendering.

For DLSS, Nvidia inputs a game into a training algorithm to determine what the visuals are supposed to look like at the sharpest possible fidelity. And this is one of the areas where DLSS 2.0 is a significant leap forward. Nvidia originally needed a bespoke training model for every game. DLSS 2.0, however, uses the same neural network for every game. This means Nvidia can add DLSS support to more games at a more rapid pace.

Then using that deep-learning data, DLSS can then use the Tensor GPU cores on Nvidias RTX cards to process what a 1080p frame should look like at 4K. And this method is so much more powerful than sharpening because it is rebuilding from data that isnt even necessarily present in each frame. Heres the result:

MechWarrior 5: Mercenaries and Control are the first two games to support DLSS 2.0. They will get the benefit of the more efficient AI network. This version of the tech is twice as fast on the Tensor cores already available in RTX cards like the RTX 2060 up to the RTX 2080 Ti.

Nvidia has also added temporal feedback to its DLSS system. This enables the super-sampling method to get information about how objects and environments change over time. DLSS 2.0 can then use that temporal feedback to improve the sharpness and stability from one frame to the next.

But the advantages go beyond improved processing. DLSS 2.0 also turns over more control to the player. One of the disadvantages of DLSS in many games is that it was often a binary choice. Either it was on or off, and developers got to decide how DLSS behaved.

DLSS 2.0 flips that by giving three presets: Quality, Balanced, and Performance. In Performance mode, DLSS 2.0 can take a 1080p frame and upscale it all the way up to 2160p (4K). Quality mode, meanwhile, may upscale 1440p to 2160p.

But you dont necessarily need a 4K display to get the advantages of DLSS 2.0. You can use the tech on a 1080p or 1440p display, and it will often provide better results than native rendering.

Again, this is possible because DLSS 2.0 is working from more data than a native 1080p frame. And all of this is going to result in higher frame rates and playable games even when using ray tracing.

DLSS 2.0 is rolling out soon as part of a driver update for RTX cards.

Read the rest here:
Nvidias DLSS 2.0 aims to prove the technology is essential - VentureBeat

Andy Renshaw, Feedzai: It crosses borders but doesnt have a passport. Its meant to protect people but can make them angry. Its competitive by nature but doesnt want you to fail. What is it?

If the PSD2 regulations and Strong Customer Authentication (SCA) feel like a riddle to you, youre not alone. SCA places strict two-factor authentication requirements upon financial institutions (FIs) at a time when FIs are facing stiff competition for customers. On top of that, the variety of payment types, along with the sheer number of transactions, continue to increase.

According to UK Finance, the number of debit card transactions surpassed cash transactions since 2017, while mobile banking surged over the past year, particularly for contactless payments. The number of contactless payment transactions per customer is growing; this increase in transactions also raises the potential for customer friction.

The number of transactions isnt the only thing thats shown an exponential increase; the speed at which FIs must process them is too. Customers expect to send, receive, and access money with the swipe of a screen. Driven by customer expectations, instant payments are gaining traction across the globe with no sign of slowing down.

Considering the sheer number of transactions combined with the need to authenticate payments in real-time, the demands placed on FIs can create a real dilemma. In this competitive environment, how can organisations reduce fraud and satisfy regulations without increasing customer friction?

For countries that fall under PSD2s regulation, the answer lies in the one known way to avoid customer friction while meeting the regulatory requirement: keep fraud rates at or below SCA exemption thresholds.

How machine learning keeps fraud rates below the exemption threshold to bypass SCA requirements

Demonstrating significantly low fraud rates allows financial institutions to bypass the SCA requirement. The logic behind this is simple: if the FIs systems can prevent fraud at such high rates, they've demonstrated their systems are secure without authentication.

SCA exemption thresholds are:

Exemption Threshold Value

Remote electronic card-based payment

Remote electronic credit transfers

EUR 500

below 0.01% fraud rate

below 0.01% fraud rate

EUR 250

below 0.06% fraud rate

below 0.01% fraud rate

EUR 100

below 0.13% fraud rate

below 0.015% fraud rate

Looking at these numbers, you might think that achieving SCA exemption thresholds is impossible. After all, bank transfer scams rose 40% in the first six months of 2019. But state-of-the-art technology rises to the challenge of increased fraud. Artificial intelligence, and more specifically machine learning, makes achieving SCA exemption thresholds possible.

How machine learning achieves SCA exemption threshold values

Every transaction has hundreds of data points, called entities. Entities include time, date, location, device, card, cardless, sender, receiver, merchant, customer age the possibilities are almost endless. When data is cleaned and connected, meaning it doesnt live in siloed systems, the power of machine learning to provide actionable insights on that data is historically unprecedented.

Robust machine learning technology uses both rules and models and learns from both historical and real-time profiles of virtually every data point or entity in a transaction. The more data we feed the machine, the better it gets at learning fraud patterns. Over time, the machine learns to accurately score transactions in less than a second without the need for customer authentication.

Machine learning creates streamlined and flexible workflows

Of course, sometimes, authentication is inevitable. For example, if a customer who generally initiates a transaction in Brighton, suddenly initiates a transaction from Mumbai without a travel note on the account, authentication should be required. But if machine learning platforms have flexible data science environments that embed authentication steps seamlessly into the transaction workflow, the experience can be as customer-centric as possible.

Streamlined workflows must extend to the fraud analysts job

Flexible workflows arent just important to instant payments theyre important to all payments. And they cant just be a back-end experience in the data science environment. Fraud analysts need flexibility in their workflows too. They're under pressure to make decisions quickly and accurately, which means they need a full view of the customer not just the transaction.

Information provided at a transactional level doesnt allow analysts to connect all the dots. In this scenario, analysts are left opening up several case managers in an attempt to piece together a complete and accurate fraud picture. Its time-consuming and ultimately costly, not to mention the wear and tear on employee satisfaction. But some machine learning risk platforms can show both authentication and fraud decisions at the customer level, ensuring analysts have a 360-degree view of the customer.

Machine learning prevents instant payments from becoming instant losses

Instant payments can provide immediate customer satisfaction, but also instant fraud losses. Scoring transactions in real-time means institutions can increase the security around the payments going through their system before its too late.

Real-time transaction scoring requires a colossal amount of processing power because it cant use batch processing, an efficient method when dealing with high volumes of data. Thats because the lag time between when a customer transacts and when a batch is processed makes this method incongruent with instant payments. Therefore, scoring transactions in real-time requires supercomputers with super processing powers. The costs associated with this make hosting systems on the cloud more practical than hosting at the FIs premises, often referred to as on prem. Of course, FIs need to consider other factors, including cybersecurity concerns before determining where they should host their machine learning platform.

Providing exceptional customer experiences by keeping fraud at or below PSD2s SCA threshold can seem like a magic trick, but its not. Its the combined intelligence of humans and machines to provide the most effective method we have today to curb and prevent fraud losses. Its how we solve the friction-security puzzle and deliver customer satisfaction while satisfying SCA.

About Andy Renshaw

Andy Renshaw, Vice President of Banking Solutions at Feedzai, has over 20 years of experience in banking and the financial services industry, leading large programs and teams in fraud management and AML. Prior to joining Feedzai, Andy held roles in global financial institutions such as Lloyds Banking Group, Citibank, and Capital One, where he helped fight against the ever-evolving financial crime landscape as a technical expert, fraud prevention expert, and a lead product owner for fraud transformation.

About Feedzai

Feedzai is the market leader in fighting fraud with AI. Were coding the future of commerce with todays most advanced risk management platform powered by big data and machine learning. Founded and developed by data scientists and aerospace engineers, Feedzai has one mission: to make banking and commerce safe. The worlds largest banks, processors, and retailers use Feedzais fraud prevention and anti-money laundering products to manage risk while improving customer experience.

Read this article:
PSD2: How machine learning reduces friction and satisfies SCA - The Paypers

Deep Learning

One of the factors often credited for this latest boom in artificial intelligence (AI) investment, research, and related cognitive technologies, is the emergence of deep learning neural networks as an evolution of machine algorithms, as well as the corresponding large volume of big data and computing power that makes deep learning a practical reality. While deep learning has been extremely popular and has shown real ability to solve many machine learning problems, deep learning is just one approach to machine learning (ML), that while having proven much capability across a wide range of problem areas, is still just one of many practical approaches. Increasingly, were starting to see news and research showing the limits of deep learning capabilities, as well as some of the downsides to the deep learning approach. So are peoples enthusiasm of AI tied to their enthusiasm of deep learning, and is deep learning really able to deliver on many of its promises?

The Origins of Deep Learning

AI researchers have struggled to understand how the brain learns from the very beginnings of the development of the field of artificial intelligence. It comes as no surprise that since the brain is primarily a collection of interconnected neurons, AI researchers sought to recreate the way the brain is structured through artificial neurons, and connections of those neurons in artificial neural networks. All the way back in 1940, Walter Pitts and Warren McCulloch built the first thresholded logic unit that was an attempt to mimic the way biological neurons worked. The Pitts and McCulloch model was just a proof of concept, but Frank Rosenblatt picked up on the idea in 1957 with the development of the Perceptron that took the concept to its logical extent. While primitive by todays standards, the Perceptron was still capable of remarkable feats - being able to recognize written numbers and letters, and even distinguish male from female faces. That was over 60 years ago!

Rosenblatt was so enthusiastic in 1959 about the Perceptrons promises that he remarked at the time that the perceptron is the embryo of an electronic computer that [we expect] will be able to walk, talk, see, write, reproduce itself and be conscious of its existence. Sound familiar? However, the enthusiasm didnt last. AI researcher Marvin Minsky noted how sensitive the perceptron was to small changes in the images, and also how easily it could be fooled. Maybe the perceptron wasnt really that smart at all. Minsky and AI researcher peer Seymour Papert basically took apart the whole perceptron idea in their Perceptrons book, and made the claim that perceptrons, and neural networks like it, are fundamentally flawed in their inability to handle certain kinds of problems notably, non-linear functions. That is to say, it was easy to train a neural network like a perceptron to put data into classifications, such as male/female, or types of numbers. For these simple neural networks, you can graph a bunch of data and draw a line and say things on one side of the line are in one category and things on the other side of the line are in a different category, thereby classifying them. But theres a whole bunch of problems where you cant draw lines like this, such as speech recognition or many forms of decision-making. These are nonlinear functions, which Minsky and Papert proved perceptrons incapable of solving.

During this period, while neural network approaches to ML settled to become an afterthought in AI, other approaches to ML were in the limelight including knowledge graphs, decision trees, genetic algorithms, similarity models, and other methods. In fact, during this period, IBMs DeepBlue purpose-built AI computer defeated Gary Kasparov in a chess match, the first computer to do so, using a brute-force alpha-beta search algorithm (so-called Good Old-Fashioned AI [GOFAI]) rather than new-fangled deep learning approaches. Yet, even this approach to learning didnt go far, as some said that this system wasnt even intelligent at all.

Yet, the neural network story doesnt end here. In 1986, AI researcher Geoff Hinton, along with David Rumelhart and Ronald Williams, published a research paper entitled Learning representations by back-propagating errors. In this paper, Hinton and crew detailed how you can use many hidden layers of neurons to get around the problems faced by perceptrons. With sufficient data and computing power, these layers can be calculated to identify specific features in the data sets they can classify on, and as a group, could learn nonlinear functions, something known as the universal approximation theorem. The approach works by backpropagating errors from higher layers of the network to lower ones (backprop), expediting training. Now, if you have enough layers, enough data to train those layers, and sufficient computing power to calculate all the interconnections, you can train a neural network to identify and classify almost anything. Researcher Yann Lecun developed LeNet-5 at AT&T Bell Labs in 1998, recognizing handwritten images on checks using an iteration of this approach known as Convolutional Neural Networks (CNNs), and researchers Yoshua Bengio and Jrgen Schmidhube further advanced the field.

Yet, just as things go in AI, research halted when these early neural networks couldnt scale. Surprisingly very little development happened until 2006, when Hinton re-emerged onto the scene with the ideas of unsupervised pre-training and deep belief nets. The idea here is to have a simple two-layer network whose parameters are trained in an unsupervised way, and then stack new layers on top of it, just training that layers parameters. Repeat for dozens, hundreds, even thousands of layers. Eventually you get a deep network with many layers that can learn and understand something complex. This is what deep learning is all about: using lots of layers of trained neural nets to learn just about anything, at least within certain constraints.

In 2010, Stanford researcher Fei-Fei Li published the release of ImageNet, a large database of millions of labeled images. The images were labeled with a hierarchy of classifications, such as animal or vehicle, down to very granular levels, such as husky or trimaran. This ImageNet database was paired with an annual competition called the Large Scale Visual Recognition Challenge (LSVRC) to see which computer vision system had the lowest number of classification and recognition errors. In 2012, Geoff Hinton, Alex Krizhevsky, and Ilya Sutskever, submitted their AlexNet entry that had almost half the number of errors as all previous winning entries. What made their approach win was that they moved from using ordinary computers with CPUs, to specialized graphical processing units (GPUs) that could train much larger models in reasonable amounts of time. They also introduced now-standard deep learning methods such as dropout to reduce a problem called overfitting (when the network is trained too tightly on the example data and cant generalize to broader data), and something called the rectified linear activation unit (ReLU) to speed training. After the success of their competition, it seems everyone took notice, and Deep Learning was off to the races.

Deep Learnings Shortcomings

The fuel that keeps the Deep Learning fires roaring is data and compute power. Specifically, large volumes of well-labeled data sets are needed to train Deep Learning networks. The more layers, the better the learning power, but to have layers you need to have data that is already well labeled to train those layers. Since deep neural networks are primarily a bunch of calculations that have to all be done at the same time, you need a lot of raw computing power, and specifically numerical computing power. Imagine youre tuning a million knobs at the same time to find the optimal combination that will make the system learn based on millions of pieces of data that are being fed into the system. This is why neural networks in the 1950s were not possible, but today they are. Today we finally have lots of data and lots of computing power to handle that data.

Deep learning is being applied successfully in a wide range of situations, such as natural language processing, computer vision, machine translation, bioinformatics, gaming, and many other applications where classification, pattern matching, and the use of this automatically tuned deep neural network approach works well. However, these same advantages have a number of disadvantages.

The most notable of these disadvantages is that since deep learning consists of many layers, each with many interconnected nodes, each configured with different weights and other parameters theres no way to inspect a deep learning network and understand how any particular decision, clustering, or classification is actually done. Its a black box, which means deep learning networks are inherently unexplainable. As many have written on the topic of Explainable AI (XAI), systems that are used to make decisions of significance need to have explainability to satisfy issues of trust, compliance, verifiability, and understandability. While DARPA and others are working on ways to possibly explain deep learning neural networks, the lack of explainability is a significant drawback for many.

The second disadvantage is that deep learning networks are really great at classification and clustering of information, but not really good at other decision-making or learning scenarios. Not every learning situation is one of classifying something in a category or grouping information together into a cluster. Sometimes you have to deduce what to do based on what youve learned before. Deduction and reasoning is not a fort of deep learning networks.

As mentioned earlier, deep learning is also very data and resource hungry. One measure of a neural networks complexity is the number of parameters that need to be learned and tuned. For deep learning neural networks, there can be hundreds of millions of parameters. Training models requires a significant amount of data to adjust these parameters. For example, a speech recognition neural net often requires terabytes of clean, labeled data to train on. The lack of a sufficient, clean, labeled data set would hinder the development of a deep neural net for that problem domain. And even if you have the data, you need to crunch on it to generate the model, which takes a significant amount of time and processing power.

Another challenge of deep learning is that the models produced are very specific to a problem domain. If its trained on a certain dataset of cats, then it will only recognize those cats and cant be used to generalize on animals or be used to identify non-cats. While this is not a problem of only deep learning approaches to machine learning, it can be particularly troublesome when factoring in the overfitting problem mentioned above. Deep learning neural nets can be so tightly constrained (fitted) to the training data that, for example, even small perturbations in the images can lead to wildly inaccurate classifications of images. There are well known examples of turtles being mis-recognized as guns or polar bears being mis-recognized as other animals due to just small changes in the image data. Clearly if youre using this network in mission critical situations, those mistakes would be significant.

Machine Learning is not (just) Deep Learning

Enterprises looking at using cognitive technologies in their business need to look at the whole picture. Machine learning is not just one approach, but rather a collection of different approaches of various different types that are applicable in different scenarios. Some machine learning algorithms are very simple, using small amounts of data and an understandable logic or deduction path thats very suitable for particular situations, while others are very complex and use lots of data and processing power to handle more complicated situations. The key thing to realize is that deep learning isnt all of machine learning, let alone AI. Even Geoff Hinton, the Einstein of deep learning is starting to rethink core elements of deep learning and its limitations.

The key for organizations is to understand which machine learning methods are most viable for which problem areas, and how to plan, develop, deploy, and manage that machine learning approach in practice. Since AI use in the enterprise is still continuing to gain adoption, especially these more advanced cognitive approaches, the best practices on how to employ cognitive technologies successfully are still maturing.

Read the original:

Are We Overly Infatuated With Deep Learning? - Forbes

John Leyden09 March 2020 at 15:50 UTC Updated: 09 March 2020 at 16:04 UTC

Forget crowdsourcing, heres crooksourcing

Computer scientists in the US are working to apply machine learning techniques in order to develop more effective honeypot-style cyber defenses.

So-called deception technology refers to traps or decoy systems that are strategically placed around networks.

These decoy systems are designed to act as a honeypot so that once an attacker has penetrated a network, they will attempt to attack them setting off security alerts in the process.

Deception technology is not a new concept. Companies including Illusive Networks and Attivo have been working in the field for several years.

Now, however, researchers from the University of Texas at Dallas (UT Dallas) are aiming to take the concept one step further.

The DeepDig (DEcEPtion DIGging) technique plants traps and decoys onto real systems before applying machine learning techniques in order to gain a deeper understanding of attackers behavior.

The technique is designed to use cyber-attacks as free sources of live training data for machine learning-based intrusion detection systems.

Somewhat ironically, the prototype technology enlists attackers as free penetration testers.

Dr Kevin Hamlen, endowed professor of computer science at UT Dallas, explained: Companies like Illusive Networks, Attivo, and many others create network topologies intended to be confusing to adversaries, making it harder for them to find real assets to attack.

The shortcoming of existing approaches, Dr Hamlen, told The Daily Swig is that such deceptions do not learn from attacks.

While the defense remains relatively static, the adversary learns over time how to distinguish honeypots from a real asset, leading to an asymmetric game that the adversary eventually wins with high probability, he said.

In contrast, DeepDig turns real assets into traps that learn from attacks using artificial intelligence and data mining.

Turning real assets into a form of honeypot has numerous advantages, according to Dr Hamlen.

Even the most skilled adversary cannot avoid interacting with the trap because the trap is within the real asset that is the adversary's target, not a separate machine or software process, he said.

This leads to a symmetric game in which the defense continually learns and gets better at stopping even the most stealthy adversaries.

The research which has applications in the field of web security was presented in a paper (PDF) entitled Improving Intrusion Detectors by Crook-Sourcing, at the recent Computer Security Applications Conference in Puerto Rico.

The research was funded by the US federal government. The algorithms and evaluation data developed so far have been publicly released to accompany the research paper.

Its hoped that the research might eventually find its way into commercially available products, but this is still some time off and the technology is still only at the prototype stage.

In practice, companies typically partner with a university that conducted the research theyre interested in to build a full product, a UT Dallas spokesman explained. Dr Hamlens project is not yet at that stage.

RELATED Gold-nuggeting: Machine learning tool simplifies target discovery for pen testers

See the article here:
AI-powered honeypots: Machine learning may help improve intrusion detection - The Daily Swig

Originally published in TechCrunch, March 16, 2020

In a briefing on Monday, research leaders across tech, academia and the government joined the White House to announce an open data set full of scientific literature on the novel coronavirus. The COVID-19 Open Research Dataset, known as CORD-19, will also add relevant new research moving forward, compiling it into one centralized hub. The new data set is machine readable, making it easily parsed for machine learning purposes a key advantage according to researchers involved in the ambitious project.

In a press conference, U.S. CTO Michael Kratsios called the new data set the most extensive collection of machine readable coronavirus literature to date. Kratsios characterized the project as a call to action for the AI community, which can employ machine learning techniques to surface unique insights in the body of data. To come up with guidance for researchers combing through the data, the National Academies of Sciences, Engineering, and Medicine collaborated with the World Health Organization to come up with high priority questions about the coronavirus related to genetics, incubation, treatment, symptoms and prevention.

The partnership, announced today by the White House Office of Science and Technology Policy, brings together the Chan Zuckerberg Initiative, Microsoft Research, the Allen Institute for Artificial Intelligence, the National Institutes of Healths National Library of Medicine, Georgetown Universitys Center for Security and Emerging Technology, Cold Spring Harbor Laboratory and the Kaggle AI platform, owned by Google.

The database brings together nearly 30,000 scientific articles about the virus known as SARS-CoV-2. as well as related viruses in the broader coronavirus group. Around half of those articles make the full text available. Critically, the database will include pre-publication research from resources like medRxiv and bioRxiv, open access archives for pre-print health sciences and biology research.

To continue reading this article, click here.

Go here to read the rest:
With Launch of COVID-19 Data Hub, The White House Issues A 'Call To Action' For AI Researchers - Machine Learning Times - machine learning & data...