Research Director of Yandex School of Data Analysis and Director of the Department of Data Analysis.
Professor at Rutgers University in New Jersey. His first work on image recognition was published in 1964, when he first proposed using integral kernels for building adaptive nonlinear classifiers. Subsequently, they became one of the main elements of all modern machine learning methods.
Currently, he is developing special methods of combinatorial optimization for the problems of cluster analysis, and uses them to solve practical problems of signal and image analysis. Has published eight monographs and 200 articles in peer-reviewed journals. He received 11 patents for the use of the developed methods in bioinformatics, medical diagnostics, individual planning of cancer treatment, practical agricultural epidemiology, regional sociogeography, and financial risk forecasting systems. Since 1967, 12 Doctors of Science have defended their dissertations under his leadership.