Category Archives: Artificial Intelligence

University of Birmingham Dubai invites applications forBSc and MSc in Artificial Intelligence and Computer Science

The University of Birmingham, Dubai, is inviting applications for its BSc and MSc courses in Artificial Intelligence and Computer Science.

Artificial intelligence (AI) is a multidisciplinary field that connects computing, psychology, neuroscience, philosophy, mathematics and linguistics. Using state-of-the-art facilities in the new, technology enhanced 'smart' campus, the degree programmes provide students with computing skills to enter industry, while also allowing them to acquire scientific skills in order to pursue research. The wide range of modules focus on areas of AI that interest students the most and will help kick-start a great career.

Both the programmes by the university have been designed in such a way that people even from a variety of academic backgrounds can get upto speed with AI and computer science without any issues.

University of Birmingham Dubai graduates are highly sought after by leading industry players, and the range of job opportunities available to them are infinite.

Programme delivery: With teaching being a joint exercise between academics on Dubai campus and Edgbaston Campus, the programmes aim to provide students with world class education. Students also have the opportunity to interact and work alongside students in the Edgbaston campus in the UK whilst they study at the Dubai campus. The BSc programme is a three-year UAE Ministry of Education accredited Computer Science programme making The University of Birmingham, the only British institution in Dubai offering three year courses. The MSc programme is the standard international course spanning across one year.

Course date:September 2022

Eligibility:

Artificial Intelligence and Computer Science BSc students with 75% from Class XII from ISC, CBSE, Maharashtra and West Bengal State Board or 80% from other state boards will be eligible to apply.

Mathematics: This course also requires students to have 85% or above in Mathematics in Class XII. If students do not meet the entry requirements for maths, they may be eligible to take a maths entry test.

Artificial Intelligence and Computer Science MScafour-year or a three-year bachelors degree in any subject other than computing with 55-60% from a recognised institution in India. Applicants can be considered on a case-by-case basis.

Application process and admission details:Applications for September 2022 are open and there is no application fee. To submit an online application, please click on the Apply Now button on the course page.

Link -https://www.birmingham.ac.uk/dubai/departments/computer-science/index.aspx

The following information will usually be needed before a decision can be made on an application:

One academic reference (or if appropriate to the programme applied for, it could be from applicants employer).

Academic transcript(s) for all prior degrees - originals or certified photocopies.

Personal Statement, approximately 5000 characters, explaining why you are interested in studying on your chosen programme.

Certificate to show competency in the English language if English is not your first language and you have already taken a test (not mandatory to submit at the time of application). Students can be considered for a waiver.

Applicants may be required to submit further documents in support of their application.

Tuition:

BSc Artificial Intelligence and Computer Science AED 120,294 (Approx. 24,162 ) per year.

MSc Artificial Intelligence and Computer Science AED 130,270 (Approx. 26,165) for 1 year full-time.

Read more here:

BSc and MSc in Artificial Intelligence and Computer Science - The Tribune

BOSTON, Feb. 18, 2022 /PRNewswire-PRWeb/ --ODSC is pleased to announce that its AI Expo & Demo Hall will be returning this April 19th-21st both in-person in Boston and virtually. Co-located with ODSC East 2022, this event is solution-focused, allowing decision-makers to learn more about what's trending in data science and artificial intelligence in 2022, AI solutions, and machine learning platforms and services. Attending this event is now free, granting attendees access to all partner booths, keynotes, and demo talks.

The AI Expo is open to all. The theme for the 2022 AI Expo & Demo Hall is Build AI Better. While many companies are already exploring the AI space, the AI Expo & Demo Hall provides a more interactive opportunity for attendees to learn how AI can improve ROI. Focus areas include machine learning, deep learning, NLP, and MLOps/data engineering. New tracks for 2022 include Responsible AI, AI Safety, AI for Biotech & Pharma.

Attendees will be able to hear from a series of keynote speakers. This year, keynotes include Padhraic Smyth of UC Irvine, Hilary Mason of Hidden Door, Luis Vargas of Microsoft, and Dr. Hari Bhaskar and Jean-Rene Gauthier of Oracle. Additionally, attendees can benefit from a number of extra events, including a networking reception, the AI Startups Showcase, the Women in Data Science Ignite session, and the AI Investors Reverse Pitch.

A key part of the AI Expo & Demo Hall is the attending vendors. This year, a few standout companies include Oracle, Microsoft Azure, Z by HP, SAS, Neo4j, iterative, Cloudera, and Iguazio. More than 40 companies will be participating this year. In the expo hall, partners will provide thought leadership and insights into current developments of data science and artificial intelligence, as well as provide examples of their most recent work. In-person and virtual offerings will differ, encouraging attendees to experience each option.

Interested attendees can register for their free ticket by going to the ODSC East 2022 AI Expo & Demo Hall page here, and experience the event across three days from April 19th-21st.

More on ODSC:

Open Data Science Conference (ODSC) is the leader of applied data science conferences. Our conferences bring industry leaders, key executives, start-up companies, engineers, and investors on the threshold of innovation together.

Media Contact

Alex Landa, ODSC, 1 8454712635, alex.l@odsc.com

SOURCE ODSC

Read the original here:

ODSC Announces Artificial Intelligence Expo is Back In-Person with Focus on MLOPs, AI Safety, and Responsible AI - WFMZ Allentown

Just what you've always wanted: A computer that falls in love with you. As refinements in artificial intelligence continue to progress the technology upward and out of the uncanny valley, getting an AI to interact with humans in a way that feels, well, human has remained one of the elusive milestones in earning it widespread practical adoption.

That all (quite literally) sounds like it's changing, and after hearing the wispy, yearning voice that AI startup Sonantic has crafted from nothing but ones and zeroes, were not sure were ready. To commemorate Valentines Day, the company shared a lengthy audio clip that shows off the emotive power of its AI-enabled voice 'bot. While it's still just working off a pre-programmed script, lets just say the gap between human and AI expression is closing fast.

Check it out:

As Sonantics cooing AI shows, were not dealing with some 1980s-vintage sci-fi cyborg that pines for human company in a cold metallic voice. The fact that it sounds so real is what weirds us out the most, in fact: If this bot were dishing out such convincing TMI on the other end of a phone call, how would you even know you werent talking to a machine?

Maybe were just hung up on love in particular, because its just one of many human emotions Sonantics AI is designed to persuasively vocalize. Speaking with The Verge, company CEO Zeena Qureshi described its algorithm tech as Photoshop for voice, capable of emulating a range of feelings from anger and fear to happiness and exuberant joy.

Though other companies are pursuing their own versions of AI voices that mimic people, co-founder John Flynn told The Verge that Sonantics technology has achieved an extra measure of user-controlled customization that helps it stand out. [O]ur research goal was to see if we could model subtle emotions. Bigger emotions are a little easier to capture, he said. I think thats the main difference our ability to direct and control and edit and sculpt a performance.

The company reportedly took inspiration for its romantic robo-voice from Samantha, the Scarlett Johansson-portrayedAI assistant in Spike Jonzes Her that main character Theordore (Joaquin Phoenix) falls in love with and for now, entertainment and marketing remain the primary fields where the tech is likely to be used. Our clients are mostly AAA game studios, entertainment studios, and were branching out into other industries, Flynn said. We recently did a partnership with Mercedes [to customize its in-car digital assistant] earlier this year.

Hey, its tough to find true love in a world that moves as fast as ours. And just like Hers Theodore, there are probably real people out there whore ready to fall for an all-too-real-sounding robot. We just hope Sonantics next-level techthe first AI that can flirt, as the company puts it doesnt end up getting its own heart broken in the process.

The rest is here:

This cutting-edge, emotional AI voice can flirt and say I love you - and sounds a little too real - Syfy

In brief

In the not too distant past, many were convinced that Artificial Intelligence (AI) or Machine Learning (ML) would not substantially change the practice of law. The legal profession was considered to be by its very nature requiring specialist skills and nuanced judgment that only humans could provide and would therefore be immune to the disruptive changes brought about by the digital transformation. However, the application of ML technology in the legal sector is now increasingly mainstream, particularly as a tool to save time for lawyers and provide a richer analysis of ever-larger datasets to aid legal decision-making in judicial systems throughout the world.

One key area of ML application in judicial systems is in predictive justice. This involves using ML algorithms that perform a probabilistic analysis of any given particular legal dispute using case law precedents. In order to work correctly, these systems must rely on huge databases of previous judicial decisions which have to be translated into a standardized language that, in turn, is able to create predetermined models. These will ultimately help the machine learning software generate the prediction.

Does this technology mean trials ending at the speed of the light, lawyers being able to know in advance whether or not to start a lawsuit, courts immediately deciding a case? Well, there is still a long way to go and we need to also balance the risks inherent to the use of these technology tools. For example, the data used to train the ML system could result in bias and consolidate stereotypes and inequality that would be validated merely because they were produced several times by the AI. Watch out, then, for possible added complexity in creating new precedents and case law against all the odds!

To assess the opportunities and challenges brought about by predictive justice systems using ML tools, it is instructive to look at case law examples, as often history is a proxy to understand the future.

The first time predictive justicestarted to see the light was in the United States way back in 2013 inState v. Loomiswhere it was used by the court in the context of sentencing. In that case, Mr. Loomis, a US citizen, was charged with driving a car in a drive-by shooting, receiving stolen goods and resisting arrest. During the trial, the circuit court was assisted in its sentencing decision by a predictive machine learning tool and the ultimate result was the judge imposing a custodial sentence. Apparently, the judge was convinced by the fact that the machine learning software tool had suggested there was a high probability that the defendant would re-offend in the same manner.

On appeal, theSupreme Court of Wisconsinaffirmed the legitimacy of the software as the judge would have reached the same result with or without the machine learning software. The decision included finding that the risk assessment provided by the AI software, although not determinative in itself, may be used as one tool to enhance a judges evaluation, weighing in the application of other sentencing evidence when deciding on appropriate sentencing for a defendant.

In essence, the Supreme Court of Wisconsin recognized the importance of the role of the judge, stating that this kind of machine learning software would not replace their role, but may be used to assist them. As we can imagine, this case opened the door to a new way of delivering justice.

Indeed, fast forward to today and we read about news from Shanghai telling us the story of the first robot ever created to analyze case files and charge defendants based on a verbal description of the case. AI scientists honed the robot using a huge amount of cases so that the machine would be able to identify various types of crimes (i.e., fraud, theft, gambling) with a claimed 97% accuracy.

AI-based predictions used to assist the courts are increasingly prevalent and may raise significant concerns (including bias and transparency). Several regulatory authorities are cooperating to advance a set of rules, principles and guidance to regulate AI platforms in judicial systems and more generally.

For example, in Europe, a significant step towards digital innovation in judicial systems was taken with the creation of the European Commission for the Efficiency of Justice (CEPEJ) which published the European Ethical Charter on the use of Artificial Intelligence in judicial systems and their environment, one of the first regulatory documents on AI (Charter). The Charter provides a set of principles to be used by legislators, law professionals and policymakers when working with AI/ML tools aimed at ensuring that the use of AI in judicial systems is compatible with the fundamental rights, including those in theEuropean Convention on Human Rightsand theConvention for the Protection of Individuals with regard to Automatic Processing of Personal Data.

Recently, CEPEJ has laid down its 2022 to2025 Action plan for the Digitalisation for a better justiceidentifying a three-step path aimed at guaranteeing a fair use of AI in courts as per the visual representation below:

Source:EUROPEAN COMMISSION FOR THE EFFICIENCY OF JUSTICE(CEPEJ) Revised roadmap for ensuring an appropriate follow-up of the CEPEJ Ethical Charter on the use of artificial intelligence in judicial systems and their environment.

CEPEJs commitment does not stop there. Indeed, the table below shows a glimpse of how information technology tools are catching on in the judicial systems of the EU Member States (civil and criminal) and how the use of IT in EU courts is accelerating.

Source:Dynamic database of European judicial systems.

More broadly, the European Commission is currently focused on developing a set of provisions to regulate AI systems which are outlined in a draftAI Regulation(Regulation) published in 2021. The Regulation proposes harmonized rules for applications of AI systems. It follows a proportionate risk-based approach differentiating among prohibited, high-risk, limited and minimal-risk uses of AI systems. Regulatory intervention, therefore, increases along with the increase in the potential of algorithmic systems to cause harm. For more, see our alertNew Draft Rules on the Use of Artificial Intelligence.

AI systems used for law enforcement or in the administration of justice are defined as high-risk AI systems under the Regulation. Note that the use of real-time biometric identification systems in public places by law enforcement is (subject to certain exceptions) prohibited. High-risk AI systems are subject to requirements, including ensuring the quality of data sets used to train the AI systems, applying human oversight, creating records to enable compliance checks and providing relevant information to users. Various stakeholders, including providers, importers, distributors and users of AI systems, are subject to individual requirements, including in relation to compliance of the AI systems with the requirements of the Regulation and CE marking of such systems to indicate conformity with the Regulation.

The Regulation has still a long way to go before being finally approved and becoming binding on Member States, but it is already a step forward in regulating AI not only as it may be used in the administration of justice, but as it may also impact deeply on the way we work, communicate, play, live in the digital era.

Camilla Ambrosinohashelped in preparing this editorial.

This article was originally published in theJanuary 2022 edition of LegalBytes.

Go here to see the original:

International: Artificial Intelligence in the administration of justice - GlobalComplianceNews

Overview

If you wonder if there is a relationship between Artificial Intelligence and cryptocurrency, then you must know that the crypto market uses Artificial Intelligence in various ways. Due to the use of AI, the crypto market can operate 24X7, and for that, the traders or investors can keep a watch on the price all the time. It creates a large amount of data for AI to examine in order to estimate future prices using back-data discoveries such as analyzing the market price. As a result, crypto AI can assist in more accurate price predictions. Because it eliminates the danger of human error while calculating, they are also faster predictions.

Nonetheless, if you want to start your crypto journey, we recommend Bitcoin Era Software. This crypto exchange platform uses advanced AI technology to target the most possible profitable crypto coins.

Moreover, if you want to know the role of artificial intelligence in the crypto market, then our blog post will be helpful for you. In the below section, we have discussed how artificial intelligence is used in cryptocurrency trading. Along with that, we have also highlighted how AI is benefiting the crypto market.

The role of artificial intelligence in crypto trading

Artificial intelligence helps the crypto market create patterns in different dimensions that can determine the pricing and volume of the crypto market. Along with that, AI can also help analyze the publics sentiment ( the market demand) or the blockchain-related data, such as speed of mining, movement of the crypto coins, the volume of transactions made by the public, and others. The AI can help you to analyze all the key aspects that are required for making profitable trading in the crypto market.

Benefits of using Artificial Intelligence

Now when you know what role AI plays in crypto trading, lets discuss the benefits of using AI. The following are some of the benefits that you get from the use of AI in the crypto market:

In a nutshell

Artificial intelligence has a significant role in the crypto market, and it can be beneficial for both the market and the public. Thus, if the crypto market starts using more advanced algorithms, then the crypto market can get more growth in the digital financial market.

Original post:

How Artificial Intelligence is used in The Crypto Market - Deadline News

The proportion of beverage companies hiring for Artificial Intelligence-related positions rose significantly last month, according to recent research, with 41.2% recruiting for at least one position in the area.

The figure, which features in GlobalDatas latest analysis, represents an increase on the 31.2% of companies recruiting for AI positions in beverages in January 2021. The rate in December last year was also lower, at 36.4%.

Of all advertised vacancies in the global beverage industry, 1.3% were linked to AI in January, a decline on the 1.6% proportion in the same month of 2021.

The analysis shows that beverage companies are currently hiring for Artificial Intelligence jobs at a lower rate compared to the average for all companies within GlobalData's job analytics database. The average among all companies stood at 2.9% for the month.

Artificial Intelligence is one of the topics that GlobalData has identified as a key disruptive force facing industry in the coming years. Companies that are investing in these areas now are considered to be better prepared for the future business landscape.

GlobalData's job analytics database tracks the daily hiring patterns of companies worldwide, drawing in vacancies as they're posted and tagging them with additional layers of data, from the seniority of each position to whether a job is linked to wider industry trends.

The themes set to shape 2022 for alcohol brand owners Click here for a Just Drinks focus

Brewing & Distilling Solutions

Global Extracts and Ingredients Manufacturer

Read this article:

Artificial Intelligence hiring levels on the rise in beverages - data - just-drinks.com

As we contemplate a future with AI in an African context, we need regulations that make sense for our continent and cultures, writes Zale Hechter.

We may think of Artificial Intelligence as some obscure, futuristic concept of the Fourth Industrial Revolution associated with scenes from "out-there" sci-fi movies likeStar WarsandTerminator. But AI is already here and being used in a wide variety of technological developments that make our daily responsibilities more efficient and convenient. In fact, scientists believe we are already 'cyborgs' because we already have a 'digital self', which lives into perpetuity.

Artificial intelligence (AI) is intelligence demonstrated by machines and contrasts with natural intelligence displayed by humans.Towards Data Sciencesays "Artificial Intelligence is the ability of a computer program to learn and think. Everything can be considered Artificial intelligence if it involves a program doing something that we would normally think would rely on the intelligence of a human."

Artificial intelligence is currently being used and employed in increasing measure across all spheres of business and industry. It is technology behind navigation and ridesharing apps you use when ordering a ride and the facial recognition functionality of digital banking. Digital assistants like Siri, Alexa and Google Assistance are all AI-powered and can take voice commands and translate them into actions. Incredibly, AI can also be used to do the chores we hate most: there is an AI-powered vacuum cleaner that scans a room, identifies things in its way and finds the quickest route to clean a space. AI is behind Netflix's personalised 'watchlists' as well as driverless vehicles.

As we contemplate a future with AI in an African context, it's interesting to consider the thoughts of prominent ex-South African, SpaceX and Tesla founder, Elon Musk. Musk is very close to the cutting edge of Artificial Intelligence, and he admits that 'it scares the hell out of me'.

READ |If machines can be inventors, could AI soon monopolise technology?

For efficiency, accuracy and agility in a digital economy and in order to have one that can trade and collaborate with other nations, AI is here to stay. Simply put, those industries and businesses that don't move with the times and fail to adopt AI are going to be left behind.

Musk highlights that AI is capable of vastly more than anyone knows, and the rate of improvement is exponential. The output of goods and services will be expedited, and there's the potential for goods and services to become more cost-effective. With automation will come abundance.

However, that abundance will come at a cost as Musk sees mass unemployment as a big social challenge of the future. He predicts that a universal basic income will be necessary as there will be fewer and fewer jobs that a robot cannot do better.

The difficulty then, of course, is how we find meaning as human beings? What would our purpose be if we are displaced by robots? How do we ensure the future is one we like and want?

Musk states that"We need to find some way of ensuring that the advent of digital super intelligence is one that is symbiotic with humanity. That's the single biggest existential crisis we face and the most pressing one".

AI in an African context

As I see it, Africa has a unique context and although current AI technology is readily available in digital banking, navigation and ride-sharing apps used on the continent, we urgently need to put regulations and policies in place that govern the use of AI in our businesses and societies.

Our greatest danger is that in our bid to keep up with the rest of the world, we leave our people without employment opportunities, having a catastrophic impact on our already fragile post-Covid-19 economic recovery.

Self-driving is rapidly developing and is becoming safer than a person driving on the road, resulting in substantial reductions in accidents. But what does the coming of autonomous vehicles in South Africa mean for our taxi industries, which are the most predominantly used mode of public transport in Africa? Would it put the industry out of business? How is our Department of Transport going to regulate their introduction to our roads? And how will the taxi industry react? Will loneliness and social disconnectedness increase if we rely on AI instead of human connections and relationships to survive and thrive?

READ |OPINION | Artificial intelligence is hijacking art history

The threat of AI to the African continent is that if we don't regulate how AI is used, it has the potential to extend and expand the digital divide by making humans redundant. Beyond that, AI can threaten the relationships that make us uniquely African.

Although part of being human and part of being alive is making mistakes, we need our modern-day workforce to strive for excellence, personalised service, and customer connection, which the best AI will never be able to replicate.

I think the job force needs to recognise that robots and technologies employing AI are, in fact, competitors to our human workforce. If bad service, serious inefficiencies, and human error cost businesses too much, they will no doubt prefer to implement AI solutions over human capital.

The importance of regulation

Musk, who is not normally an advocate for regulation, argues that regulation and oversight are critically important when it comes to AI. He argues this is a case where a very serious danger exists to the public and calls for public bodies that have insight and oversight to ensure AI is being developed safely. The danger of AI is much greater than nuclear warheads, he says. The regulator oversight is frighteningly lacking.

The importance of retaining our humanity

As I see it, AI should not strip us of the core of what makes us human.

The implementation and adoption of AI in an African context needs to take place in a symbiotic way that does not erode or diminish our African essence. Culturally in Africa, relationships and communities and the expression of our cultures and religion are incredibly important. We need to ensure that our rich heritage, culture and diversity are not destroyed or diminished by our desire for convenience and efficiency.

Likewise, AI should be developed with the mindset of uplifting people and building economies that support - and does not compete with - humanity.

- Zale Hechtor is theCEO of Cliqtech the company that created SmartWill.

To receive Opinions Weekly, sign up for the newsletterhere.

*Want to respond to the columnist? Send your letter or article toopinions@news24.comwith your name and town or province. You are welcome to also send a profile picture. We encourage a diversity of voices and views in our readers' submissions and reserve the right not to publish any and all submissions received.

Disclaimer:News24 encourages freedom of speech and the expression of diverse views. The views of columnists published on News24 are therefore their own and do not necessarily represent the views of News24.

See the original post here:

OPINION | Zale Hechter: Artificial intelligence is here to stay, and it will affect our future - News24

Dataset

Kvasir is a multi-class dataset from Brum Hospital in Vestre Viken Health Trust (Norway), collected from 2010 to 201424. Kvasir (v2) contains 8000 endoscopic images labelled with eight distinct classes, with approximately 1000 images per class, including ulcerative colitis. The images are assigned only image-level labels, provided by at least one experienced endoscopist as well as medical trainees (minimum of 3 reviewers per label). The images are independent, with only one image per patient.

Standard endoscopy equipment was used. HyperKvasir is an extension of the Kvasir dataset, collected from the same Brum Hospital from 2008 to 2016, containing 110,079 images, 10,662 of which are labelled with 23 classes of findings25. Pathological findings in particular accounted for 12 of 23 classes, which are aggregated and summarized in Table 1. They can be broadly grouped into Barrets esophagus and esophagitis in the upper GI tract, and polyps, ulcerative colitis, and hemorrhoids in the lower GI tract.

Importantly, the dataset includes 851 ulcerative colitis images which are labelled and graded using the Mayo endoscopic subscore26,27 by a minimum of one board certified gastroenterologist and one or more junior doctors or PhD students (total of 3 reviewers per image). The images are in JPEG format, with varying image resolutions, the most common being 576768, 576720, and 10721920. Table 2 shows the number of images available for each Mayo grade.

The HyperKvasir study, including the HyperKvasir dataset available through the Center for Open Science we are using here, was approved by Norwegian Privacy Data Protection Authority, and exempted from patient consent because the data were fully anonymous. All metadata was removed, and all files renamed to randomly generated file names before the internal IT department at Brum hospital exported the files from a central server. The study was exempted from approval from the Regional Committee for Medical and Health Research EthicsSoutheast Norway since the collection of the data did not interfere with the care given to the patient. Since the data is anonymous, the dataset is publicly shareable and complies with Norwegian and General Data Protection Regulation (GDPR) laws. Apart from this, the data has not been pre-processed or augmented in any way.

Two binary classification tasks were formulated from the dataset:

Diagnosis: All pathological findings of ulcerative colitis were grouped along with all other classes of pathological findings in the dataset (Fig.1a). The problem was formulated as a binary classification task to distinguish UC from non-UC pathology on endoscopic still images.

Methods (a) Overview of methods used to train diagnostic classification model of ulcerative colitis from multi-class endoscopic images on Kvasir datasets. (b) Overview of methods used to train diagnostic model for endoscopic grading of ulcerative colitis on HyperKvasir dataset.

Grading: Evaluation of disease severity using endoscopic images of UC pathology. Mayo graded image labels were binned into Grades 01 and 23. (Fig.1b) This grouping has been used in previous machine learning studies and for clinical trial endpoints19. Therefore, the task was to distinguish inactive/mild from moderate/severe UC.

A filter was designed to remove the green picture-in-picture depicting the endoscope. The filter applied a uniform crop to all images, filling in the missing pixels with 0 values, turning them black.

Source images were then normalized to [1, 1] and downscaled to 299299 resolution using bilinear resampling. Images underwent random transformations of rotation, zoom, sheer, vertical and horizontal flip, using a set seed. Image augmentation was only applied to training set images (not validation or test set), inside each fold of the fivefold cross-validation.

There are a growing variety of machine learning frameworks that could provide the foundation for our study. Our choices here acknowledge the current dominance of deep neural network methods, despite the emerging challenges of explainability (explainable artificial intelligence=XAI) and trust in practical clinical implementation41. Most of our choices use the most popular method for classifying images (convolutional neural networks), whose major differences lie in their depth of layering (50160) and recorded dimensionality of annotated relationships amongst segments of images (up to 2048).

The following four different CNN architectures were tested on the Kvasir dataset:

Pre-trained InceptionV3, a 159-layer CNN. The output of InceptionV3 in this configuration is a 2048-dimensional feature vector28.

Pre-trained ResNet50, a Keras implementation of ResNet50, a 50-layer CNN which uses residual functions that reference previous layer inputs29.

Pre-trained VGG19, a Keras implementation of VGG which is a 19 layer CNN developed by Visual Geometry Group30.

Pre-trained DenseNet121, a Keras implementation of DenseNet with 121 layers31.

All pre-trained models were TensorFlow implementations initialized using ImageNet weights32.Training was performed end-to-end with no freezing of layers. All models performed a final classification step via a dense layer with one node. Sigmoid activation was used at this final dense layer, with binary cross entropy for the models loss function.

For both classification tasks, the final dataset was randomly shuffled and split into training and validation sets in a 4:1 ratio, where 80% images were used for fivefold cross-validation and 20% unseen images were used for evaluating model performance. The best model from each fold were combined and used as the final model for prediction on the test set.

Hyperparameters were fine-tuned using Grid Search, where the search space included the following parameters: optimizers: Adam, Stochastic Gradient Descent (SGD), learning rate: 0.01, 0.001, 0.0001; momentum (for SGD): 0, 0.5, 0.9, 0.99. For all models, training phases consisted of 20 epochs with batch size of 32.

Models were evaluated using accuracy, recall, precision, and F1-scores. As a binary classification problem, confusion matrices and ROC curves were used to visualize model performance.

To provide visual explanation of what the models are learning, we chose the Gradient-weighted Class Activation Mapping (Grad-CAM) technique33. Grad-CAM produces a heatmap for each model output, showing which part(s) of the image the model is using to make predictions (produces the strongest activation). The heatmap is a course localization map produced by using gradient information flowing into the last convolutional neural network layer, to assign importance values to each neuron.

We also had an experienced gastroenterologist (D.C.B.) annotate and highlight the regions of interest in representative images to provide a comparison with the regions of interest generated by the heatmaps.

Model building was performed and figures created was done using TensorFlow and Keras packages32 in Python 3.6.9, run on Google Colab (https://research.google.com/colaboratory/) notebook.

Go here to read the rest:

Artificial intelligence enabled automated diagnosis and grading of ulcerative colitis endoscopy images | Scientific Reports - Nature.com

Artificial Intelligence

Human-Centered Artificial Intelligence (HCAI) is a concept that seems to put human usage and access of AI technology at the forefront. To me, it seems in opposition to the data driven vision of some pundits, though there is the ability to differentiate between goals and development. Human-Centered AI, by Ben Shneiderman, is an excellent introduction to the concepts of HCAI. Be aware, though, that this isnt a breezy, short, book aimed at quick review. This is a business school textbook. For management interested in governance and control, focus on part four of the book, discussed towards the end of this article. The section should be a must-read, even if you skim the rest.

That point is important so as not to surprise people. The books audience should be business personnel and students wanting a strong introduction to the issues of HCAI showing concepts that should then be drilled down into practice. It is for upper- and middle-management in the CIO, CTO, R&D and other more technical realms of an organization. The text is 376 pages in a font smaller than the usual business book. Give that content, this review will remain at a higher level than many of the book reviews in this column.

Theres an important thread running through the book. The author differentiates two different research lenses that can be used, that of science and innovation. The science approach is focused on what is possible from a technical view. Why it is being done doesnt matter. On the other hand, Ben Shneiderman points to the innovation view, that of understanding how a technology can provide innovation in the real world. HCAI is driven from the innovation perspective.

As much as I like this book, it isnt perfect. The big problem early is in chapter four, and that chapter should be skimmed or skipped. In it, the author presents another academic technologists view that the AI revolution is similar to the industrial revolution and makes the same mistake many do in claiming jobs wont be lost. The industrial revolution took people from farms and crafts into simple shop floors. Over generations, those shop floors became more complex, but it was a stepwise advancement. Artificial intelligence isnt that. It will take over jobs with no similar positions to fill. The gap between many of the disappearing jobs and the remaining ones are much larger than during the industrial revolution.

He also states that automation lowers cost and improves quality. The first, yes. The second is very arguable. That, however, is a discussion for another day.

Back to what I like. Chapter eight focuses on the authors two dimensional view of human and automation controls, how they will overlap. There are some excellent examples.

Chapter 12 is a key for understanding the science v innovation views mentioned above. While the discussion runs through the book, this chapter focuses on it in a clear way. It also discusses why the innovation view requires understanding and explainability of AI systems.

Social robots are described and discussed in detail in chapter 16. While it is a good survey of options, I do think the author misses one critical point. He points out that surveys over the years show people interested in anthropomorphic robots, with the feedback implying those robots arent yet good enough. Then he states, in softer words, the opinion that they will never be good enough. Too many opinions over the years have stated because AI hasnt yet reached point X, theyll never reach point X. Thats a stretch.

The same chapter points to what the author describes as supertools, functional devices as the alternative. They avoid the anthropomorphic trap to create usable devices with responses that are accepted. They are useful, and clearly a segment of devices that will remain; but chatbot research has also shown improving technology, creating more acceptance as long as people know they are talking with a chatbot.

For business managers and government employees who wish to better understand the organizational impact of AI at multiple levels, part four is the meat of the matter. The author defines four levels of governance:

Software development

Corporate policy

Industry & trade standards

Governmental regulations

While the entire book is a good overview of HCAI, much of it is aimed at a mid-tier management, and even development manager, level of focus. The explanation of the four levels is something that is important to all levels in companies, industry and government. The types of governance arent independent, and people must be aware of how they integrate. For instance, if software developers arent paying attention to social demands, they wont be prepared for government that could lay waste to expenditures of time and money. In the opposite direction, there is not much use of creating industry standards and government regulations to arent directly applicable to the technology.

That means government officials hiring people to better educate them in whats possible. Long before the recent hearings on social media, and even before the famous statement by Senator Ted Stevens about the internet being as series of tubes, legislators have wanted to help citizens but not understood the implications of the technology and how best to address it.

It also means that corporate policy is critical, as it must be the bridge between development and the real world. Human-centered AI is an important concept. This book is a heavy introduction, and many parts of it will be useful to different audiences. Students, in academia and business, can read it all, but it is still valuable to management who need to both understand how to better direct AI development and to require appropriate AI to solve market and social challenges.

Originally posted here:

Human-Centered AI, An In-Depth Study Of The Current State Of The Artificial Intelligence Concept - Forbes

Artificial intelligence is a transformative technology that impacts every aspect of businesses globally in key business decisions. Most US businesses are faced with the daunting challenge of understanding AI and its adoption. As a result, AI and analytics professionals are in ever-increasing demand as the technology is integrated into business systems.

To understand why AI is becoming so popular, the impact of AI and the field of AI and data science as a career choice, the University of Louisville College of Business is organising a webinar on 23rd of February 2022 from 7:30 to 8:30 PM to discuss ways in which one can manage their career in the age of AI.

Jeff Guan is a Professor of Computer Information Systems in the College of Business, University of Louisville. He holds a doctoral degree in computer engineering and is the creator and director of the Master of Science in Business Analytics program at UofL. His research interests include accounting information systems, machine learning, knowledge management, and information systems implementation and adoption.

He has published and presented about 100 papers in the above areas and teaches courses in database, artificial intelligence, data warehousing, and information systems analysis and designboth at the undergraduate and graduate levels. He also has extensive consulting experience with private and government organisations, such as Brown-Forman, GE Appliances, Yum!, Department of Defense, and Kentucky State Government.

Swan is an experienced international education professional with demonstrated leadership in international student services, international programming, campus internationalisation, and international student recruitment. He will walk the participants through the graduate admission process, discuss internship opportunities, and offer answers to any questions they might have.

View original post here:

Managing your career in the age of artificial intelligence: Register for this webinar - Analytics India Magazine