Kanani, N. Electroplating: Basic Principles, Processes and Practice 1st edition (Elsevier, 2004).

Singh, J., Srivastava, A. N., Singh, N. & Singh, A. Stability Constants of Metal Complexes in Solution. in Stability and Applications of Coordination Compounds (ed. Srivastava, A. N.) (IntechOpen, 2019).

Treybal, R. E. Mass transfer Operations (Springer, 1980).

Bruijnincx, P. C. A. & Sadler, P. J. New trends for metal complexes with anticancer activity. Curr. Opin. Chem. Biol. 12, 197206 (2008).

CAS PubMed PubMed Central Article Google Scholar

Dimmock, P. W., Warwick, P. & Robbins, R. A. Approaches to predicting stability constants. Analyst 120, 21592170 (1995).

ADS CAS Article Google Scholar

Janet, J. P. & Kulik, H. J. Predicting electronic structure properties of transition metal complexes with neural networks. Chem. Sci. 8, 51375152 (2017).

CAS PubMed PubMed Central Article Google Scholar

Meyer, B., Sawatlon, B., Heinen, S., Anatole von Lilienfeld, O. & Corminboeuf, C. Machine learning meets volcano plots: computational discovery of cross-coupling catalysts. Chem. Sci. 9, 70697077 (2018).

CAS PubMed PubMed Central Article Google Scholar

Chaube, S., Goverapet Srinivasan, S. & Rai, B. Applied machine learning for predicting the lanthanide-ligand binding affinities. Sci. Rep. 10, 14322 (2020).

ADS CAS PubMed PubMed Central Article Google Scholar

Solovev, V., Kireeva, N., Ovchinnikova, S. & Tsivadze, A. The complexation of metal ions with various organic ligands in water prediction of stability constants by QSPR ensemble modelling. J. Incl. Phenom. Macrocycl. Chem. 83, 89101 (2015).

Article CAS Google Scholar

Tetko, I. V., Solovev, V. P. & Antonov, A. V. Benchmarking of linear and nonlinear approaches for quantitative structure-property relationship studies of metal complexation with ionophores. J. Chem. Inf. Model. 46, 808819 (2006).

CAS PubMed Article Google Scholar

Solovev, V., Marcou, G., Tsivadze, A. & Varnek, A. Complexation of Mn2+, Fe2+, Y3+, La3+, Pb2+, and UO22+ with organic ligands: QSPR ensemble modeling of stability constants. Ind. Eng. Chem. Res. 51, 1348213489 (2012).

Article CAS Google Scholar

Solovev, V. P., Tsivadze, A. Y. & Varnek, A. A. New approach for accurate QSPR modeling of metal complexation: Application to stability constants of complexes of lanthanide ions Ln3+ Ag+, Zn2+, Cd2+ and Hg2+ with organic ligands in water. Macroheterocycles 5, 404410 (2012).

Article CAS Google Scholar

Solvev, V. P., Kireeva, N., Tsivadze, Y. & Varnek, A. QSPR ensemble modelling of alkaline-earth metal complexation. J. Incl. Phenom. Macrocycl. Chem. 76, 159171 (2013).

Article CAS Google Scholar

Solvev, V. et al. Stability constants of complexes of Zn2+, Cd2+, and Hg2+ with organic ligands: QSPR consensus modeling and design of new metal binders. J. Incl. Phenom. Macrocycl. Chem. 72, 309321 (2012).

Article CAS Google Scholar

Baskin, I. I., Solovev, V. P., Bagaturyants, A. A. & Varnek, A. Predictive cartography of metal binders using generative topographic mapping. J. Comput. Aided. Mol. Des. 31, 701714 (2017).

ADS CAS PubMed Article Google Scholar

Quang, N. M., Nhung, N. T. A. & Tat, P. V. An insight QSPR-based prediction model for stability constants of metal-thiosemicarbazone complexes using MLR and ANN methods. Vietnam J. Chem. 57, 500506 (2019).

Article CAS Google Scholar

Shiri, F., Salahinejad, M., Momeni-Mooguei, N. & Sanchooli, M. Predicting stability constants of transition metals; Y3+, La3+, and UO22+ with organic ligands using the 3D-QSPR methodology. J. Recept. Signal Transduct. Res. 41, 5966 (2021).

CAS PubMed Article Google Scholar

Solovev, V., Varnek, A. & Tsivadze, A. QSPR ensemble modelling of the 1:1 and 1:2 complexation of Co2+, Ni2+, and Cu2+ with organic ligands: relationships between stability constants. J. Comput. Aided. Mol. Des. 28, 549564 (2014).

ADS PubMed Article CAS Google Scholar

Deringer, V. L. et al. Gaussian process regression for materials and molecules. Chem. Rev. 121, 1007310141 (2021).

CAS PubMed PubMed Central Article Google Scholar

Motoyama, Y. et al. Bayesian optimization package: PHYSBO. Comput. Phys. Commun. 278, 108405 (2022).

MathSciNet CAS Article Google Scholar

Zhang, H. Inconsistent estimation and asymptotically equal interpolations in model-based geostatistics. J. Am. Stat. Assoc. 99, 250261 (2004).

MathSciNet MATH Article Google Scholar

Piironen, J. & Vehtari, A. Projection predictive model selection for Gaussian processes. 2016 IEEE 26th International Workshop on Machine Learning for Signal Processing (MLSP), 2016, 16 (2016).

Paananen, T., Piironen, J., Andersen, M. R. & Vehtari, A. Variable selection for Gaussian processes via sensitivity analysis of the posterior predictive distribution. Proc. 22nd Int Conf. Artig. Intell. Statist. 89, 17431752 (2019).

Google Scholar

Gasteiger, J. & Marsili, M. Iterative partial equalization of orbital electronegativity-A rapid access to atomic charges. Tetrahedron 36, 32193228 (1980).

CAS Article Google Scholar

Hall, L. H. & Kier, L. B. Electrotopological state indices for atom types: A novel combination of electronic, topological, and valence state information. J. Chem. Inf. Comput. Sci. 35, 10391045 (1995).

CAS Article Google Scholar

Smith, R. M. & Martell, A. E. NIST Critically Selected Stability Constants of Metal Complexes Database (NIST Standard Reference Database 46). version 8.0, (National Institute of Science and Technology, Gaithersburg, MD, 2004). https://www.nist.gov/srd/nist46. Accessed 1 March 2022.

Fernandez-Botello, A., Griesser, R., Hol, A., Moreno, V. & Sigel, H. Acidbase and metal-ion-binding properties of 9-[2-(2-Phosphonoethoxy)ethyl]adenine (PEEA), a relative of the antiviral nucleotide analogue 9-[2-(Phosphonomethoxy)ethyl]adenine (PMEA). An exercise on the quantification of isomeric complex equilibria in solution. Inorg. Chem. 44, 51045117 (2005).

CAS PubMed Article Google Scholar

Kapinos, L. E., Hol, A., Gnter, J. & Sigel, H. Metal ion-binding properties of 1-Methyl-4-aminobenzimidazole (=9-Methyl-1,3-dideazaadenine) and 1,4-Dimethylbenzimidazole (=6,9-Dimethyl-1,3-dideazapurine). Quantification of the steric effect of the 6-Amino group on metal ion binding at the N7 site of the adenine residue. Inorg. Chem. 40, 25002508 (2001).

CAS PubMed Article Google Scholar

Melton, D. L., VanDerveer, D. G. & Hancock, R. D. Complexes of greatly enhanced thermodynamic stability and metal ion size-based selectivity, formed by the highly preorganized non-macrocyclic ligand 1,10-Phenanthroline-2,9-dicarboxylic Acid. A thermodynamic and crystallographic study. Inorg. Chem. 45, 93069314 (2006).

CAS PubMed Article Google Scholar

Sigel, H., Da Costa, C. P., Song, B., Carloni, P. & Greg, F. Stability and structure of metal ion complexes formed in solution with acetyl phosphate and acetonylphosphonate: Quantification of isomeric equilibria. J. Am. Chem. Soc. 121, 62486257 (1999).

CAS Article Google Scholar

Klmn, F. K. et al. Synthesis, Potentiometric, Kinetic, and NMR Studies of 1,4,7,10-Tetraazacyclododecane-1,7-bis(acetic acid)-4,10-bis(methylenephosphonic acid) (DO2A2P) and its Complexes with Ca(II), Cu(II), Zn(II) and Lanthanide(III) Ions. Inorg. Chem. 47, 38513862 (2008).

PubMed Article CAS Google Scholar

Nonat, A., Gateau, C., Fries, P. H. & Mazzanti, M. Lanthanide complexes of a picolinate ligand derived from 1,4,7-Triazacyclononane with potential application in magnetic resonance imaging and time-resolved luminescence imaging. Chem. Eur. J. 12, 71337150 (2006).

CAS PubMed Article Google Scholar

Kotek, J. et al. Study of thermodynamic and kinetic stability of transition metal and lanthanide complexes of DTPA analogues with a phosphorus acid pendant arm. Eur. J. Inorg. Chem. 2006, 19761986 (2006).

Article CAS Google Scholar

Rodrguez, L. et al. Anion detection by fluorescent Zn(II) complexes of functionalized polyamine ligands. Inorg. Chem. 47, 61736183 (2008).

PubMed Article CAS Google Scholar

Aragoni, M. C. et al. Coordination chemistry of N-aminopropyl pendant arm derivatives of mixed N/S-, and N/S/O-donor macrocycles, and construction of selective fluorimetric chemosensors for heavy metal ions. Dalton Trans. 2005, 29943004 (2005).

Article CAS Google Scholar

Caltagirone, C. et al. Redox chemosensors: coordination chemistry towards CuII, ZnII, CdII, HgII, and PbII of 1-aza-4,10-dithia-7-oxacyclododecane ([12]aneNS2O) and its N-ferrocenylmethyl derivative. Dalton Trans. 2003, 901909 (2003).

Article CAS Google Scholar

Bazzicalupi, C. et al. Protonation and coordination properties towards Zn(II), Cd(II) and Hg(II) of a phenanthroline-containing macrocycle with an ethylamino pendant arm. Dalton Trans. 2004, 591597 (2004).

Article CAS Google Scholar

Blake, A. J. et al. A new pyridine-based 12-membered macrocycle functionalised with different fluorescent subunits; coordination chemistry towards CuII, ZnII, CdII, HgII, and PbII. Dalton Trans. 2004, 27712779 (2004).

Article CAS Google Scholar

Baranyai, Z., Bombieri, G., Meneghetti, F., Tei, L. & Botta, M. A solution thermodynamic study of the Cu(II) and Zn(II) complexes of EBTA: X-ray crystal structure of the dimeric complex [Cu2(EBTA)(H2O)3]2. Inorg. Chim. Acta 362, 22592264 (2009).

CAS Article Google Scholar

Miguirditchian, M. et al. Thermodynamic Study of the Complexation of Trivalent Actinide and Lanthanide Cations by ADPTZ, a Tridentate N-Donor Ligand. Inorg. Chem. 44, 14041412 (2005).

CAS PubMed Article Google Scholar

Kobayashi, T. et al. Effect of the introduction of amide oxygen into 1,10-Phenanthroline on the extraction and complexation of trivalent lanthanide in acidic condition. Sep. Sci. Technol. 45, 24312436 (2010).

CAS Article Google Scholar

Miguirditchian, M. et al. Complexation of Lanthanide(III) and Actinide(III) cations with tridentate nitrogen-donor ligands: A luminescence and spectrophotometric study. Nucl. Sci. Eng. 153, 223232 (2006).

CAS Article Google Scholar

Ogden, M. D., Sinkov, S. I., Meier, G. P., Lumetta, G. J. & Nash, K. L. Complexation of N4-Tetradentate ligands with Nd(III) and Am(III). J. Solut. Chem. 41, 21382153 (2012).

CAS Article Google Scholar

Merrill, D. & Hancock, R. D. Metal ion selectivities of the highly preorganized tetradentate ligand 1,10-phenanthroline-2,9-dicarboxamide with lanthanide(III) ions and some actinide ions. Radiochim. Acta 99, 161166 (2011).

CAS Article Google Scholar

Reddy, K. H., Prasad, N. B. L. & Reddy, T. S. Analytical properties of 1-phenyl-1,2-propanedione-2-oxime thiosemicarbazone: simultaneous spectrophotometric determination of copper(II) and nickel(II) in edible oils and seeds. Talanta 59, 425433 (2003).

CAS PubMed Article Google Scholar

Veeranna, V., Rao, V. S., Laxmi, V. V. & Varalankshmi, T. R. Simultaneous second order derivative spectrophotometric determination of cadmium and cobalt using furfuraldehyde Thiosemicarbazone (FFTSC). Res. J. Phyarm. Tech. 6, 577584 (2013).

Google Scholar

Atalay, T. & zkan, E. Evaluation of thermodynamic parameters and stability constants of Cu(II), Ag(I) and Hg(II) complexes of 2-methylindole-3-carboxaldehyde thiosemicarbazone. Thermochim. Acta 244, 291295 (1994).

CAS Article Google Scholar

Sharma, S. R. K. & Sindhwani, S. K. Thermal studies on the chelation behavior of biologically active 2-hydroxy-1-naphthaldehyde thiosemicarbazone (HNATS) towards bivalent metal ions: A potentiometric study. Thermochim. Acta 202, 291299 (1992).

Article Google Scholar

Draho, B. et al. Mn2+ complexes with 12-membered pyridine based macrocycles bearing carboxylate or phosphonate pendant arm: Crystallographic, thermodynamic, kinetic, redox, and 1H/17O relaxation studies. Inorg. Chem. 50, 1278512801 (2011).

PubMed Article CAS Google Scholar

Draho, B., Kotek, J., Hermann, P., Luke, I. & Toth, . Mn2+ Complexes with pyridine-containing 15-membered macrocycles: thermodynamic, kinetic, crystallographic, and 1H/17O relaxation studies. Inorg. Chem. 49, 32243238 (2010).

PubMed Article CAS Google Scholar

Svobodov, I. et al. Thermodynamic, kinetic and solid-state study of divalent metal complexes of 1,4,8,11-tetraazacyclotetradecane (cyclam) bearing two trans (1,8-)methylphosphonic acid pendant arms. Dalton Trans. 2006, 51845197 (2006).

Article Google Scholar

Bazzicalupi, C. et al. Basicity and coordination properties of a new phenanthroline-based bis-macrocyclic receptor. Dalton Trans. 2006, 40004010 (2006).

Article CAS Google Scholar

Yamada, H., Hayashi, H. & Yasui, T. Utility of 1-Octanol/Octane mixed solvents for the solvent extraction of Aluminum(III), Gallium(III), and Indium(III) with 8-Quinolinol. Anal. Sci. 22, 371376 (2006).

Original post:
Machine learning-based analysis of overall stability constants of metalligand complexes | Scientific Reports - Nature.com

Advanced computer programs influence, and can even dictate, meaningful parts of our lives. Think of streaming services, credit scores, facial recognition software.

As this technology becomes more sophisticated and more pervasive, its important to understand the basic terminology.

People often use algorithm, machine learning and artificial intelligence interchangeably. There is some overlap, but theyre not the same things.

We decided to call up a few experts to help us get a firm grasp on these concepts, starting with a basic definition of algorithm. The following is an edited transcript of the episode.

Melanie Mitchell, Davis professor of complexity at the Santa Fe Institute, offered a simple explanation of a computer algorithm.

An algorithm is a set of steps for solving a problem or accomplishing a goal, she said.

The next step up is machine learning, which uses algorithms.

Rather than a person programming in the rules, the system itself has learned, Mitchell said.

For example, speech recognition software, which uses data to learn which sounds combine to become words and sentences. And this kind of machine learning is a key component of artificial intelligence.

Artificial intelligence is basically capabilities of computers to mimic human cognitive functions, said Anjana Susarla, who teaches responsible AI at Michigan State Universitys Broad College of Business.

She said we should think of AI as an umbrella term.

AI is much more broader, all-encompassing, compared to only machine learning or algorithms, Susarla said.

Thats why you might hear AI as a loose description for a range of things that show some level of intelligence. Like software that examines the photos on your phone to sort out the ones with cats to advanced spelunking robots that explore caves.

Heres another way to think of the differences among these tools: cooking.

Bethany Edmunds, professor and director of computing programs at Northeastern University, compares it to cooking.

She says an algorithm is basically a recipe step-by-step instructions on how to prepare something to solve the problem of being hungry.

If you took the machine learning approach, you would show a computer the ingredients you have and what you want for the end result. Lets say, a cake.

So maybe it would take every combination of every type of food and put them all together to try and replicate the cake that was provided for it, she said.

AI would turn the whole problem of being hungry over to the computer program, determining or even buying ingredients, choosing a recipe or creating a new one. Just like a human would.

So why do these distinctions matter? Well, for one thing, these tools sometimes produce results with biased outcomes.

Its really important to be able to articulate what those concerns are, Edmunds said. So that you can really dissect where the problem is and how we go about solving it.

Because algorithms, machine learning and AI are pretty much baked into our lives at this point.

Columbia Universitys engineering school has a further explanation of artificial intelligence and machine learning, and it lists other tools besides machine learning that can be part of AI. Like deep learning, neural networks, computer vision and natural language processing.

Over at the Massachusetts Institute of Technology, they point out that machine learning and AI are often used interchangeably because these days, most AI includes some amount of machine learning. A piece from MITs Sloan School of Management also gets into the different subcategories of machine learning. Supervised, unsupervised and reinforcement, like trial and error with kind of digital rewards. For example, teaching an autonomous vehicle to drive by letting the system know when it made the right decision like not hitting a pedestrian, for instance.

That piece also points to a 2020 survey from Deloitte, which found that 67% of companies are already using machine learning, and 97% were planning to in the future.

IBM has a helpful graphic to explain the relationship among AI, machine learning, neural networks and deep learning, presenting them as Russian nesting dolls with the broad category of AI as the biggest one.

And finally, with so many businesses using these tools, the Federal Trade Commission has a blog laying out some of the consumer risks associated with AI and the agencys expectations of how companies should deploy it.

Continue reading here:
The ABCs of AI, algorithms and machine learning - Marketplace

All the experiments were conducted at the Agri-Food and Biosciences Institute (AFBI) farm at Hillsborough, County Down, UK. All the experiments and procedures complied with the requirements of the UK Animals (Scientific Procedures) Act 1986 and were approved by the AFBI Hillsborough Ethical Review Group. All the experiments were performed in accordance with relevant guidelines and regulations (following the ARRIVE guidelines26).

Data used were collated from 43 total diet digestibility studies with 951 lactating dairy cows undertaken at Agri-Food and Biosciences Institute in Northern Ireland over a period of 26years (19902015). The data from studies undertaken between 1990 and 2002 were used as the training dataset (n=564) and undertaken between 2005 and 2015 as the testing dataset (n=387). The training data were used to develop prediction models for MN using MLR and the three selected machine learning algorithms (ANN, RFR and SVR). These new models were then tested for their predictive performance using the training dataset by tenfold cross validation. The testing dataset were used for the independent evaluation and comparison of predictive ability of different modeling approaches. The information of the two datasets on numbers of experiments, cow genotypes and forage types offered are presented in Table 10. Data on live weight, milk production, feed intake, N intake and outputs are presented in Table 11. The datasets used in the present study showed a various cow genetic merit and a broad range in LW (379781kg), MY (5.140.2kg/d), total dry matter intake (7.5426.6kg/d), FP (0.211.00%), DNC (19.038.0g/kg DM), diet metabolizable energy concentration (DMEC, 9.6819.4MJ/kg DM) and NI (155874g/d), which represents typical dairy production conditions managed within grassland-based dairy systems in the West and North Europe.

Cows were housed in free-stall cubicle accommodation for at least 20 d before commencing digestibility trials in metabolism units for 8 d with feed intake, milk production and feces and urine collected during the final 6 d. Throughout the whole experiment, cows were offered experimental diets ad libitum and had free access to water. During the final 6 d, the following measurements for each individual cows were carried out to generate total digestibility data used in the present study. Forages and concentrates offered and refused were recorded daily and sampled for analysis of feed dry matter (DM), N concentration and forage proportion. Feces and urine outputs were collected daily and sampled for DM (feces only) and N concentration. Milk yield was recorded daily and sampled for analysis fat, protein and lactose concentrations. Live weight was measured on the first and last days in the metabolism unit. Details in feed intake, feces and urine collection and methods used for analysis of feed, feces, urine and milk samples were described by Yan et al.6.

Because features (variables) in raw data may have different dynamic ranges, which may result in poor model performance, it is recommended to normalize them to make ANN training more efficient by performing normalization process for the raw inputs10. In the present study, all the input data for ANN models were normalized into the interval [0, 1] by performing MinMax normalization technique27 using Eq.(1):

$$X_{norm} = frac{{X {-} X_{min} }}{{X_{max} - X_{min} }}$$

(1)

where Xnorm or X is the normalized or original value, Xmin or Xmax is the minimum or maximum values of the input data.

After finding the optimal tuning parameter, all normalized data for MN obtained by ANN models were denormalized into their original scale using Eq.(2) 27:

$$Y = Y_{norm} * , left( {Y_{max} - , Y_{min} } right) , + , Y_{min}$$

(2)

where Ynorm or Y is the normalized or demoralized value, Ymin or Ymax is the minimum or maximum values of the output data.

Feature selection is an essential step during development of models, which can hugely impact the generalization and predictive ability of models10,28. In the present study, a hybrid knowledge-based and data driven approach was developed and implemented to selecting features. Knowledge in animal science and the process of digestibility trial were applied to diagnosing and removing irrelevant features before the implementing of data driven feature selection process. For instance, the features of feces N output (FN) and urine N output (UN) were excluded from the set of features in present study according to prior background and expert knowledge. Because the data of UN and FN were obtained from analyzing urine and feces samples and then they were summed up and treated as new feature MN, both FN and UN are heavily correlated with MN. Their inclusion in the features list might cause poor generalization performance of the models. Furthermore, the optimal features selected from data driven approach may need to be diagnosed based on background knowledge in animal science according to the scenarios of model application. For instance, several variables (e.g. NI and FP) included in datasets used in this study may not be available in commercial farms. Therefore, alternative feature (concentrate dry matte intake, CDMI) was selected and included into the feature list in this study based on the domain knowledge and then new ANN model suits for commercial farms was developed.

The filter method was applied for feature selection using the Pearson correlation matrix and variance inflation factor (VIF) technique. The first step was to use the Pearson correlation matrix to identify features which might correlate each other for prediction of MN excretion, because using correlated features in models could influence performance of these models with a biased outcome. If two features were heavily correlated, the less important one was removed from the set of features to minimize adverse effects on model performance. Afterwards, the VIF analysis was applied to detect multicollinearity, which has been widely used as a measure of the degree of multicollinearity among input features. A VIF score was calculated for each feature and those with high values were removed. The threshold score for the VIF analysis was 5 and features with a VIF score below this threshold were selected. The VIF score was computed by VIF function in R29.

In the present study, four models based on the MLR ANN, RFR and SVR were developed using the training dataset and these new models were tested using the testing dataset for comparison of their prediction performance for MN outputs in lactating dairy cows (presented later). The MLR with the stepwise procedure for selection of independent variables was used as benchmark model since it is a well-known technique and has been applied for modelling in a wide range of applications in animal science research. Alternative modeling approaches proposed in the present study were ANN, RFR and SVR. To compare the performance, models developed with different approaches and ensure that the same resampling sets were used between calls, the same random number seeds were set prior to perform the process of training, fitting and testing models. All statistical analyses were performed with R29.

The MLR model (Eq.3) selected in the present study for the prediction of MN output was published in 20066 which was developed using the same training dataset listed in Table 2. To improve the estimation of the regression parameters, experiment was included as a random factor during the development of MLR model. The dataset had a large range within each dependent or independent variable, e.g., MN, NI, LW, MY, FP and DNC, which is vital to ensure the development of robust regression model applicable under various farming conditions10.

$${text{MN }}left( {{text{g}}/{text{d}}} right) , = , 0.{text{749 NI }} + , 0.0{text{65 LW }}{-}{ 1}.{text{515 MY }}{-}{ 17}.0$$

(3)

where NI, LW and MY are N intake (g/d), live weight (kg) and milk yield (kg/d), respectively.

In the present study, ANN was fitted using R package neuralnet which was built to train neural networks in the context of regression analyses. The details of ANN training and application of neuralnet were described by Gnther and Fritsch30. Multilayer perceptron networks trained with backpropagation learning algorithms were used and consist of an input layer, hidden layer(s) and an output layer. The input variables were obtained by using the feature selection algorithm described in the section Knowledge-based and data driven feature selection, and the neuron in output layer represents MN. The ANN models were trained based on the selection of training algorithms and learning parameters including the number of hidden layer(s), number of neurons in hidden layer(s), error function, threshold for partial derivatives of the error function as stopping criteria, and activation function etc.. The optimized number of hidden layer(s), number of neuron(s) in the hidden layer(s), learning algorithms, learning rate and other learning parameters were obtained on the basis of prediction performance measured as relative root mean square error (RRMSE, Eq.6) with tenfold cross validation and then the best topology/architecture was finalized.

The RFR is an ensemble machine learning method and a nonparametric technique derived from classification and regression trees which are constructed using a bootstrap aggregating (bagging) method from the training data31. In RFR, prediction is conducted by averaging the individual tree predictions. A detailed description of RFR theory can be found in the report by Breiman32. The RFR was implemented by the randomForest function in the R package (version 3.6.1). To select the optimal hyperparameters for learning algorithm, tuning process was performed based on the R package ranger. The hyperparameters include number of trees to grow (ntree), number of randomly drawn candidate variables (mtry), sample size and node size. Grid search strategy was used to choose the candidate hyperparameter values and the performances of the trained algorithm with different values of the hyperparameters were evaluated as RRMSE (Eq.6) by using tenfold cross validation.

The SVR uses similar principles as support vector machine, a supervised non-parametrical statistical learning technique that uses the kernel functions and the maximum margin algorithm to solve the nonlinear problem33. The detailed theoretical background and description of SVR can be found in the report by Cristianini and Shawe-Taylor34. The SVR model performs the regression estimation by risk minimization where the risk is measured by a loss function. In this study, R package e1071 was used and the svm function was implemented to fit SVR model. The radial basis kernels function, the most commonly used kernels types, was employed in training and predicting process. Parameter tuning was performed by using grid search over supplied parameter ranges and the best combination of parameters (lowest RMSE) were selected. The performance of SVR model was measured as RRMSE (Eq.6) with tenfold cross validation.

The MLR model and the three new models (ANN, RFR and SVR) was developed and compared in terms of their prediction performance for MN outputs in lactating dairy cows based on the datasets listed in Table 2. The predictive performance of models were evaluated using coefficient of determination (R2), root mean square error (RMSE), relative root mean square error (RRMSE) and concordance correlation coefficient (CCC), based on the actual and predicted values. The R2 was calculated using Eq.(4). The RMSE and RRMSE were produced in a tenfold cross validation process (10 RMSE data generated) using Eq.(5)35 and Eq.(6)36, respectively. The concordance correlation coefficient (CCC), a further measure of the agreement between observed and predicted values, was given by Eq.(7)37. The tenfold cross validation was used to evaluate prediction performance of these models (MLR, ANN, RFR and SVR)The obtained RMSE, RRMSE and CCC values (n=10) through the tenfold cross validation were compared among the 4 models using one-way analysis of variance and then followed by Tukeys honest significant difference (HSD) test (=0.05). The same cross validation folds were used for all modeling scenarios to compare cross all of the models performance.

$$R^{2} = 1 - frac{{sum left( {y_{i} - hat{y}} right)^{2} }}{{sum left( {y_{i} - overline{y}} right)^{2} }}$$

(4)

$$RMSE = sqrt { frac{1}{n}mathop sum limits_{i = 1}^{n} left( {y_{i} - hat{y}} right)^{2} }$$

(5)

$$RRMSE = (RMSE/overline{y}) times , 100$$

(6)

$$CCC = frac{{2 cdot ,r cdot ,S_{{widehat{y}}} cdot,S_{y} }}{{S_{{widehat{y}}}^{2} + S_{y}^{2} + left( {mathop sum nolimits_{i = 1}^{n} frac{{left( {y_{i} - widehat{y}} right)}}{n}} right)^{2} }}$$

(7)

where (y_{i}) is actual MN, (widehat{{y_{i} }}) is predicted MN, (overline{y}) is the mean of actual MN and n is the number of observations, r is the Pearson correlation coefficient between (widehat{{y_{i} }}) and (overline{y}), (S_{{hat{y}}}) and (S_{y}) are the respective standard divisions.

See the article here:
Can machine learning algorithms perform better than multiple linear regression in predicting nitrogen excretion from lactating dairy cows | Scientific...

Scientists have developed artificial intelligence software that can create proteins that may be useful as vaccines, cancer treatments, or even tools for pulling carbon pollution out of the air.

This research, reported today in the journal Science, was led by the University of Washington School of Medicine and Harvard University. The article is titled"Scaffolding protein functional sites using deep learning."

The proteins we find in nature are amazing molecules, but designed proteins can do so much more, said senior author David Baker, an HHMI Investigator and professor of biochemistry at UW Medicine. In this work, we show that machine learning can be used to design proteins with a wide variety of functions.

For decades, scientists have used computers to try to engineer proteins. Some proteins, such as antibodies and synthetic binding proteins, have been adapted into medicines to combat COVID-19. Others, such as enzymes, aid in industrial manufacturing. But a single protein molecule often contains thousands of bonded atoms; even with specialized scientific software, they are difficult to study and engineer.

Inspired by how machine learning algorithms can generate stories or even images from prompts, the team set out to build similar software for designing new proteins. The idea is the same: neural networks can be trained to see patterns in data. Once trained, you can give it a prompt and see if it can generate an elegant solution. Often the results are compelling or even beautiful, said lead author Joseph Watson, a postdoctoral scholar at UW Medicine.

The team trained multiple neural networks using information from the Protein Data Bank, which is a public repository of hundreds of thousands of protein structures from across all kingdoms of life. The neural networks that resulted have surprised even the scientists who created them.

The team developed two approaches for designing proteins with new functions. The first, dubbed hallucination is akin to DALL-E or other generative A.I. tools that produce new output based on simple prompts. The second, dubbed inpainting, is analogous to the autocomplete feature found in modern search bars and email clients.

Most people can come up with new images of cats or write a paragraph from a prompt if asked, but with protein design, the human brain cannot do what computers now can, said lead author Jue Wang, a postdoctoral scholar at UW Medicine. Humans just cannot imagine what the solution might look like, but we have set up machines that do.

To explain how the neural networks hallucinate a new protein, the team compares it to how it might write a book: You start with a random assortment of words total gibberish. Then you impose a requirement such as that in the opening paragraph, it needs to be a dark and stormy night. Then the computer will change the words one at a time and ask itself Does this make my story make more sense? If it does, it keeps the changes until a complete story is written, explains Wang.

Both books and proteins can be understood as long sequences of letters. In the case of proteins, each letter corresponds to a chemical building block called an amino acid. Beginning with a random chain of amino acids, the software mutates the sequence over and over until a final sequence that encodes the desired function is generated. These final amino acid sequences encode proteins that can then be manufactured and studied in the laboratory.

The team also showed that neural networks can fill in missing pieces of a protein structure in only a few seconds. Such software could aid in the development of new medicines.

With autocomplete, or Protein Inpainting, we start with the key features we want to see in a new protein, then let the software come up with the rest. Those features can be known binding motifs or even enzyme active sites, explains Watson.

Laboratory testing revealed that many proteins generated through hallucination and inpainting functioned as intended. This included novel proteins that can bind metals as well as those that bind the anti-cancer receptor PD-1.

The new neural networks can generate several different kinds of proteins in as little as one second. Some include potential vaccines for the deadly respiratory syncytial virus,orRSV.

All vaccines work by presenting a piece of a pathogen to the immune system. Scientists often know which piece would work best, but creating a vaccine that achieves a desired molecular shape can be challenging. Using the new neural networks, the team prompted a computer to create new proteins that included the necessary pathogen fragment as part of their final structure. The software was free to create any supporting structures around the key fragment, yielding several potential vaccines with diverse molecular shapes.

When tested in the lab, the team found that known antibodies against RSV stuck to three of their hallucinated proteins. This confirms that the new proteins adopted their intended shapes and suggests they may be viable vaccine candidates that could prompt the body to generate its own highly specific antibodies. Additional testing, including in animals, is still needed.

I started working on the vaccine stuff just as a way to test our new methods, but in the middle of working on the project, my two-year-old son got infected by RSV and spent an evening in the ER to have his lungs cleared. It made me realize that even the test problems we were working on were actually quite meaningful, said Wang.

These are very powerful new approaches, but there is still much room for improvement, said Baker, who was a recipient of the 2021 Breakthrough Prize in Life Sciences. Designing high activity enzymes, for example, is still very challenging. But every month our methods just keep getting better! Deep learning transformed protein structure prediction in the past two years, we are now in the midst of a similar transformation of protein design.

This project was led by Jue Wang, Doug Tischer, and Joseph L. Watson, who are postdoctoral scholars at UW Medicine, as well as Sidney Lisanza and David Juergens, who are graduate students at UW Medicine. Senior authors include Sergey Ovchinnikov, a John Harvard Distinguished Science Fellow at Harvard University, and David Baker, professor of biochemistry at UW Medicine.

Compute resources for this work were donated by Microsoft and Amazon Web Services.

Funding was provided by the Audacious Project at the Institute for Protein Design; Microsoft; Eric and Wendy Schmidt by recommendation of the Schmidt Futures; the DARPA Synergistic Discovery and Design project (HR001117S0003 contract FA8750-17-C-0219); the DARPA Harnessing Enzymatic Activity for Lifesaving Remedies project (HR001120S0052 contract HR0011-21-2-0012); the Washington Research Foundation; the Open Philanthropy Project Improving Protein Design Fund; Amgen; the Human Frontier Science Program Cross Disciplinary Fellowship (LT000395/2020-C) and EMBO Non-Stipendiary Fellowship (ALTF 1047-2019); the EMBO Fellowship (ALTF 191-2021); the European Molecular Biology Organization (ALTF 139-2018); the la Caixa Foundation; the National Institute of Allergy and Infectious Diseases (HHSN272201700059C), the National Institutes ofHealth (DP5OD026389); the National Science Foundation (MCB 2032259); the Howard Hughes Medical Institute, the National Institute on Aging (5U19AG065156); the National Cancer Institute (R01CA240339); the Swiss National Science Foundation; the Swiss National Center of Competence for Molecular Systems Engineering; the Swiss National Center of Competence in Chemical Biology; and the European Research Council(716058).

Written by Ian Haydon, UW Medicine Institute for Protein Design

View post:
Biologists train AI to generate medicines and vaccines - UW Medicine Newsroom

In this "On the Job" segment from Cheddar News, Amin Venjara, General Manager of Data Solutions at ADP, describes the importance of data and how human resources leaders are relying on real-time access to data now more than ever. Venjara offers real-world examples of data's impact on the top challenges faced by organizations today.

Businesses big and small have been utilizing the latest tech and innovation to make the new remote and hybrid working environments possible.

Speaking with Cheddar News, above, Amin Venjara (AV), says relying on quality and accessible data to take action is how today's HR teams are impacting the modern workforce.

Q: How does data influence the role of human resources (HR)?

AV: The last few years have thrust HR teams into the spotlight. Think about all the changes we've seen managing the onset of the pandemic, the return to the workplace, the great resignation and all the challenges that's brought and even the increased focus on diversity, equity and inclusion. HR has been at the focal point of responding to these challenges. And in response, we've seen an uptick in the use of workforce analytics and benchmarking. HR teams need the data to be able to help make decisions in real time as things are changing. And they're using it with the executives and managers they support to make data-driven decisions.

Q: Clearly data-driven solutions are critical in today's workforce as you've been discussing, where has data made the most significant impact?

AV: When we talk to employers, we continuously hear about four key areas related to their workforce: attracting top talent, retaining and engaging talent, optimizing labor costs, and fostering a diverse, equitable and inclusive workforce.

To give an example of the kind of impact that data can have. We have a product that helps organizations calculate and take action on pay equity. They can see gaps by gender and race ethnicity and based on internal and market data. Over 70% of active clients using this tool are seeing a decrease in pay equity gaps. If you look at the size of this - they're spending over a billion dollars to close those gaps. That's not just analytics and data - that's taking action. So, think about the impact that has on the message about equal pay for equal work. And also, the impact it has on productivity, and the lives of those individual workers and their families.

Q: In today's tight talent market, employers increasingly need help recruiting and even retaining workers. How can data and machine learning alleviate some of those very pressing challenges?

AV: Here's an interesting thing about what's happening in the current labor market. U.S. employment numbers are back to pre-pandemic levels with 150 million workers on the payroll. However, we're at the lowest unemployed workers to jobs openings rate we've seen in over 15 years. To put it simply, it's a candidate's market out there; and jobs are chasing workers.

Two things to keep in mind: employers have to employ data-driven strategies to be competitive. So we're seeing with labor markets changing, remote work, hybrid work, expectations on pay and even the physical locations of workers people have moved a lot. Employers need access to real-time data, accurate data on supply and demand of labor and on compensation to hire the right workers and keep the ones they have.

The second thing is really about the adoption of machine learning in recruiting workflows. We're seeing machine learning being adopted in chatbots for personalizing the experience and even helping with scheduling, but also AI-based algorithms to help score candidate profiles against jobs. Overall, the best organizations are combing technology and data with their recruiting and hiring managers to decrease the overall time to fill open jobs.

Q: Becoming data confident might be a concern or even perhaps intimidating for some, but what's an example of how an organization can use data well?

AV: A lot of organizations are trying to make this happen. We recently worked with a quick service restaurant with about 2,000 locations across the U.S. In light of the supply chain challenges and demographic shifts of the last couple of years, they wanted to know how to combine and optimize the supply at each location based on expected demand.

Their research enabled them to correlate demographics, things like age, income and even family status to items on the menu like salads, sandwiches and kids' meals. But what they needed was a stronger signal on what's happening in the local context of each location. They had used internal data for so long, but things had shifted. By using our monthly anonymized and aggregated data from nearly 20% of the workforce, they were able to optimize their demand forecasting models and increase their supply chain efficiency. There are two lessons to think about. They had a key strategic problem, and they worked backwards from that. That's a key piece of becoming data confident - focusing on something that matters and making a data-driven decisions about it. The second is about going beyond the four walls of your organization. There are so many different and new sources of data available due to the digitization of our economy. In order to lock the insight and the strength of signal you need you really need to look for the best sources to get there.

Q: How do you see the role of data evolving as we look toward the future of work?

AV: Data has really come the language of business right now. I see a couple of trends as we look out. The first is the acceleration of data in the flow of work. When you look at a lot of organizations today, when people need data, they have to go to a reporting group or a business intelligence group to request the data. Then it takes a couple cycles to get it right and then make a decision. The cycle time can be high.

What I expect to see now is data more and more in the flow of work where business decision makers are working immediately; they have the right data at their fingertips. You see that across domains. Second is just the separation between haves and have nots. With the increasing speed of change, data haves are going to be able to outstrip data have nots. Those who have invested in building the right organizational, technical, and cultural muscle will see the spoils of this in the years to come.

Learn more

In the post-pandemic world of work, the organizations that prioritize people first will rise to the top. Find out how to make HR more personalized to adapt to today's changing talent landscape. Get our guide: Work is personal

Tags: Compensation Diversity and Inclusion Trends and Innovation Salary and Wages Technology HCM Technology HR Recruiting and Hiring Articles

Read the original post:
How Data Has Changed the World of HR - ADP

We are excited to bring Transform 2022 back in-person July 19 and virtually July 20 - 28. Join AI and data leaders for insightful talks and exciting networking opportunities. Register today!

The sheer number of backlogs and delays across the public sector are unsettling for an industry designed to serve constituents. Making the news last summer was the four-month wait period to receive passports, up substantially from the pre-pandemic norm of 6-8 weeks turnaround time. Most recently, the Internal Revenue Service (IRS) announced it entered the 2022 tax season with 15 times the usual amount of filing backlogs, alongside its plan for moving forward.

These frequently publicized backlogs dont exist due to a lack of effort. The sector has made strides with technological advancements over the last decade. Yet, legacy technology and outdated processes still plague some of our nations most prominent departments. Todays agencies must adopt digital transformation efforts designed to reduce data backlogs, improve citizen response times and drive better agency outcomes.

By embracing machine learning (ML) solutions and incorporating advancements in natural language processing (NLP), backlogs can be a thing of the past.

Whether tax documents or passport applications, processing items manually takes time and is prone to errors on the sending and receiving sides. For example, a sender may mistakenly check an incorrect box or the receiver may interpret the number 5 as the letter S. This creates unforeseen processing delays or, worse, inaccurate outcomes.

But managing the growing government document and data backlog problem is not as simple and clean-cut as uploading information to processing systems. The sheer number of documents and citizens information entering agencies in varied unstructured data formats and states, often with poor readability, make it nearly impossible to reliably and efficiently extract data for downstream decision-making.

Embracing artificial intelligence (AI) and machine learning in daily government operations, just as other industries have done in recent years, can provide the intelligence, agility and edge needed to streamline processes and enable end-to-end automation of document-centric processes.

Government agencies must understand that real change and lasting success will not come with quick patchworks built upon legacy optical character recognition (OCR) or alternative automation solutions, given the vast amount of inbound data.

Bridging the physical and digital worlds can be attained with intelligent document processing (IDP), which leverages proprietary ML models and human intelligence to classify and convert complex, human-readable document formats. PDFs, images, emails and scanned forms can all be converted into structured, machine-readable information using IDP. It does so with greater accuracy and efficiency than legacy alternatives or manual approaches.

In the case of the IRS, inundated with millions of documents such as 1099 forms and individuals W-2s, sophisticated ML models and IDP can automatically identify the digitized document, extract printed and handwritten text, and structure it into a machine-readable format. This automated approach speeds up processing times, incorporates human support where needed and is highly effective and accurate.

Alongside automation and IDP, introducing ML and NLP technologies can significantly support the sectors quest to improve processes and reduce backlogs. NLP isan area of computer science that processes and understands text and spoken words like humans do, traditionally grounded in computational linguistics, statistics and data science.

The field has experienced significant advancements, like the introduction of complex language models that contain more than 100 billion parameters. These models could power many complex text processing tasks, such as classification, speech recognition and machine translation. These advancements could support even greater data extraction in a world overrun by documents.

Looking ahead, NLP is on course to reach the level of text understanding capability similar to that of a human knowledge worker, thanks to technological advancements driven by deep learning.Similar advancements in deep learning also enable the computer to understand and process other human-readable content such as images.

For the public sector specifically, this could be images included in disability claims or other forms or applications consisting of more than just text. These advancements could also improve downstream stages of public sector processes, such as ML-powered decision-making for agencies determining unemployment assistance, Medicaid insurance and other invaluable government services.

Though weve seen a handful of promising digital transformation improvements, the call for systemic change has yet to be fully answered.

Ensuring agencies go beyond patching and investing in various legacy systems is needed to move forward today. Patchwork and investments in outdated processes fail to support new use cases, are fragile to change and cannot handle unexpected surges in volume. Instead, introducing a flexible solution that can take the most complex, difficult-to-read documents from input to outcome should be a no-brainer.

Why? Citizens deserve more out of the agencies who serve them.

CF Su is VP of machine learning at Hyperscience.

Welcome to the VentureBeat community!

DataDecisionMakers is where experts, including the technical people doing data work, can share data-related insights and innovation.

If you want to read about cutting-edge ideas and up-to-date information, best practices, and the future of data and data tech, join us at DataDecisionMakers.

You might even considercontributing an articleof your own!

Read More From DataDecisionMakers

Read this article:
The imperative need for machine learning in the public sector - VentureBeat

Engineering | News releases | Research | Technology

July 11, 2022

ClearBuds use a novel microphone system and are one of the first machine-learning systems to operate in real time and run on a smartphone.Raymond Smith/University of Washington

As meetings shifted online during the COVID-19 lockdown, many people found that chattering roommates, garbage trucks and other loud sounds disrupted important conversations.

This experience inspired three University of Washington researchers, who were roommates during the pandemic, to develop better earbuds. To enhance the speakers voice and reduce background noise, ClearBuds use a novel microphone system and one of the first machine-learning systems to operate in real time and run on a smartphone.

The researchers presented this project June 30 at the ACM International Conference on Mobile Systems, Applications, and Services.

ClearBuds differentiate themselves from other wireless earbuds in two key ways, said co-lead author Maruchi Kim, a doctoral student in the Paul G. Allen School of Computer Science & Engineering. First, ClearBuds use a dual microphone array. Microphones in each earbud create two synchronized audio streams that provide information and allow us to spatially separate sounds coming from different directions with higher resolution. Second, the lightweight neural network further enhances the speakers voice.

While most commercial earbuds also have microphones on each earbud, only one earbud is actively sending audio to a phone at a time. With ClearBuds, each earbud sends a stream of audio to the phone. The researchers designed Bluetooth networking protocols to allow these streams to be synchronized within 70 microseconds of each other.

The teams neural network algorithm runs on the phone to process the audio streams. First it suppresses any non-voice sounds. And then it isolates and enhances any noise thats coming in at the same time from both earbuds the speakers voice.

Because the speakers voice is close by and approximately equidistant from the two earbuds, the neural network can be trained to focus on just their speech and eliminate background sounds, including other voices, said co-lead author Ishan Chatterjee, a doctoral student in the Allen School. This method is quite similar to how your own ears work. They use the time difference between sounds coming to your left and right ears to determine from which direction a sound came from.

Shown here, the ClearBuds hardware (round disk) in front of the 3D printed earbud enclosures.Raymond Smith/University of Washington

When the researchers compared ClearBuds with Apple AirPods Pro, ClearBuds performed better, achieving a higher signal-to-distortion ratio across all tests.

Its extraordinary when you consider the fact that our neural network has to run in less than 20 milliseconds on an iPhone that has a fraction of the computing power compared to a large commercial graphics card, which is typically used to run neural networks, said co-lead author Vivek Jayaram, a doctoral student in the Allen School. Thats part of the challenge we had to address in this paper: How do we take a traditional neural network and reduce its size while preserving the quality of the output?

The team also tested ClearBuds in the wild, by recording eight people reading from Project Gutenberg in noisy environments, such as a coffee shop or on a busy street. The researchers then had 37 people rate 10- to 60-second clips of these recordings. Participants rated clips that were processed through ClearBuds neural network as having the best noise suppression and the best overall listening experience.

One limitation of ClearBuds is that people have to wear both earbuds to get the noise suppression experience, the researchers said.

But the real-time communication system developed here can be useful for a variety of other applications, the team said, including smart-home speakers, tracking robot locations or search and rescue missions.

The team is currently working on making the neural network algorithms even more efficient so that they can run on the earbuds.

Additional co-authors are Ira Kemelmacher-Shlizerman, an associate professor in the Allen School; Shwetak Patel, a professor in both the Allen School and the electrical and computer engineering department; and Shyam Gollakota and Steven Seitz, both professors in the Allen School. This research was funded by The National Science Foundation and the University of Washingtons Reality Lab.

For more information, contact the team at clearbuds@cs.washington.edu.

Read more:
ClearBuds: First wireless earbuds that clear up calls using deep learning - University of Washington

We are excited to bring Transform 2022 back in-person July 19 and virtually July 20 - 28. Join AI and data leaders for insightful talks and exciting networking opportunities. Register today!

The era of Big Data has helped democratize information, creating a wealth of data and growing revenues at technology-based companies. But for all this intelligence, were not getting the level of insight from the field of machine learning that one might expect, as many companies struggle to make machine learning (ML) projects actionable and useful. A successful AI/ML program doesnt start with a big team of data scientists. It starts with strong data infrastructure. Data needs to be accessible across systems and ready for analysis so data scientists can quickly draw comparisons and deliver business results, and the data needs to be reliable, which points to the challenge many companies face when starting a data science program.

The problem is that many companies jump feet first into data science, hire expensive data scientists, and then discover they dont have the tools or infrastructure data scientists need to succeed. Highly-paid researchers end up spending time categorizing, validating and preparing data instead of searching for insights. This infrastructure work is important, but also misses the opportunity for data scientists to utilize their most useful skills in a way that adds the most value.

When leaders evaluate the reasons for success or failure of a data science project (and 87% of projects never make it to production) they often discover their company tried to jump ahead to the results without building a foundation of reliable data. If they dont have that solid foundation, data engineers can spend up to 44% of their time maintaining data pipelines with changes to APIs or data structures. Creating an automated process of integrating data can give engineers time back, and ensure companies have all the data they need for accurate machine learning. This also helps cut costs and maximize efficiency as companies build their data science capabilities.

Machine learning is finicky if there are gaps in the data, or it isnt formatted properly, machine learning either fails to function, or worse, gives inaccurate results.

When companies get into a position of uncertainty about their data, most organizations ask the data science team to manually label the data set as part of supervised machine learning, but this is a time-intensive process that brings additional risks to the project. Worse, when the training examples are trimmed too far because of data issues, theres the chance that the narrow scope will mean the ML model can only tell us what we already know.

The solution is to ensure the team can draw from a comprehensive, central store of data, encompassing a wide variety of sources and providing a shared understanding of the data. This improves the potential ROI from the ML models by providing more consistent data to work with. A data science program can only evolve if its based on reliable, consistent data, and an understanding of the confidence bar for results.

One of the biggest challenges to a successful data science program is balancing the volume and value of the data when making a prediction. A social media company that analyzes billions of interactions each day can use the large volume of relatively low-value actions (e.g. someone swiping up or sharing an article) to make reliable predictions. If an organization is trying to identify which customers are likely to renew a contract at the end of the year, then its likely working with smaller data sets with large consequences. Since it could take a year to find out if the recommended actions resulted in success, this creates massive limitations for a data science program.

In these situations, companies need to break down internal data silos to combine all the data they have to drive the best recommendations. This may include zero-party information captured with gated content, first-party website data, and data from customer interactions with the product, along with successful outcomes, support tickets, customer satisfaction surveys, even unstructured data like user feedback. All of these sources of data contain clues if a customer will renew their contract. By combining data silos across business groups, metrics can be standardized, and theres enough depth and breadth to create confident predictions.

To avoid the trap of diminishing confidence and returns from an ML/AI program, companies can take the following steps.

By building the right infrastructure for data science, companies can see whats important for the business, and where the blind spots are. Doing the groundwork first can deliver solid ROI, but more importantly, it will set up the data science team up for significant impact. Getting a budget for a flashy data science program is relatively easy, but remember, the majority of such projects fail. Its not as easy to get budget for the boring infrastructure tasks, but data management creates the foundation for data scientists to deliver the most meaningful impact on the business.

AlexanderLovell is head of product atFivetran.

Welcome to the VentureBeat community!

DataDecisionMakers is where experts, including the technical people doing data work, can share data-related insights and innovation.

If you want to read about cutting-edge ideas and up-to-date information, best practices, and the future of data and data tech, join us at DataDecisionMakers.

You might even considercontributing an articleof your own!

Read More From DataDecisionMakers

Original post:
How to make the most of your AI/ML investments: Start with your data infrastructure - VentureBeat

A research team led by Professor Sanguk Kim (Department of Life Sciences) at POSTECH is gaining attention as they have improved the accuracy of predicting patient response to immune checkpoint inhibitors (ICIs) by using network-based machine learning.

The research team discovered new network-based biomarkers by analyzing the clinical results of more than 700 patients with three different cancers (melanoma, gastric cancer, and bladder cancer) and the transcriptome data of the patients' cancer tissues. By utilizing the network-based biomarkers, the team successfully developed artificial intelligence that could predict the response to anticancer treatment.

The team further proved that the treatment response prediction based on the newly discovered biomarkers was superior to that based on conventional anticancer treatment biomarkers including immunotherapy targets and tumor microenvironment markers.

In their previous study, the research team had developed machine learning that could predict drug responses to chemotherapy in patients with gastric or bladder cancer. This study has shown that artificial intelligence using the interactions between genes in a biological network could successfully predict the patient response to not only chemotherapy, but also immunotherapy in multiple cancer types.

This study helps detect patients who will respond to immunotherapy in advance and establish treatment plans, resulting in customized precision medicine with more patients to benefit from cancer treatments. Supported by the POSTECH Medical Device Innovation Center, the Graduate School of Artificial Intelligence, and ImmunoBiome Inc, this study was recently published inNature Communications, an international peer-reviewed journal.

See the original post:
Using Machine Learning to Predict Immunotherapy Response - Medical Device and Diagnostics Industry

REDWOOD CITY, Calif.--(BUSINESS WIRE)--C3 AI (NYSE: AI), the Enterprise AI application software company, today announced that Forrester Research has named it a Leader in AI and Machine Learning Platforms in its July 2022 report, The Forrester Wave: AI/ML Platforms, Q3 2022.

Ahead of its time, C3 AIs strategy is to make AI application-centric by building a growing library of industry solutions, forging deep industry partnerships, running in every cloud, and facilitating extreme reuse through common data models, the report states.

We are pleased to be recognized as a leader in AI and ML platforms," said Thomas Siebel, C3 AI CEO. Im delighted to see C3 AIs significant investments in enterprise AI software be acknowledged. I believe that Forrester Research has made an important contribution, having published the first professional comprehensive analysis of enterprise AI and Machine Learning platforms, Siebel continued, changing the dialogue from a focus on disjointed tools to the importance of cohesive enterprise AI platforms. This is certain to accelerate the market adoption of enterprise AI and simplify often protracted decision processes.

Of the 15 vendors in the report, C3 AI received the top ranking in the Strategy category. For the following criteria, C3 AI received:

Download The Forrester Wave: AI and Machine Learning Platforms, Q3 2022 report here.

About C3 AI

C3 AI is the Enterprise AI application software company. C3 AI delivers a family of fully integrated products including the C3 AI Suite, an end-to-end platform for developing, deploying, and operating enterprise AI applications and C3 AI Applications, a portfolio of industry-specific SaaS enterprise AI applications that enable the digital transformation of organizations globally.

Read the rest here:
C3 AI Named a Leader in AI and Machine Learning Platforms - Business Wire