\u201cUnder the hood of a car, you will find an engine, under the hood of artificial intelligence, you will find machine learning\u201d<\/p>\n
The machine learning workgroup is the leading research formation in Estonia dealing with AI based diagnostic support for neurodegenerative diseases, cognitive deviations, and fatigue assessment. In this area, we focus on the analysis of drawing and writing tests that have been used in psychology, neurology, and psychiatry for more than a century. The rapidly aging population and the shortage of medical professionals in developed countries are the main reasons to devote resources to this area. The analysis of cognitive deviations aims to support the proper development and early education of schoolchildren. Lastly, fatigue is one of the factors that endanger workplace safety and cause financial losses. Although in first view diseases, cognitive deviations and fatigue may have little in common, their underlying mechanisms and techniques to recognize them are very similar. In this area, the group is cooperating with\u00a0Prof. Aaro Toomela\u00a0from Tallinn University. This cooperation has already resulted in several defended thesis, research grands, and numerous research articles in top tier journals and conference proceedings.<\/p>\n
On the international level, we cooperate with the\u00a0Analysis and PDE center of Ghent University in Belgium. Medicine and neurology, especially, are very delicate areas of research. Besides the obvious restrictions related to ethics and privacy, in many cases one must deal with very small data sets or interpret unexpected human behaviors. The latest provided us with unique expertise not only in human movement analysis, but also in synthetic data generation, the fundamentals of feature engineering and selection, and the interpretation of AI decisions.<\/p>\n
The group also cooperates with the\u00a0Department of Marine Systems\u00a0to apply state of the art machine learning and deep learning techniques to estimate Baltic sea surface conditions based on satellite born synthetic aperture radar. Another long last cooperation is with the\u00a0Cybersecurity Research Group\u00a0where the focus is to model event reported in social media and answer the questions about its lifespan in the information field, its truthiness and properties. Previously, the head of the group cooperated to solve network security problems.<\/p>\n
On the theoretical side, the group actively works on the problems of explainable AI. In this area our cooperation with the\u00a0Nonlinear Control Systems Group\u00a0is aimed at developing novel explainers and methods of their validation.<\/p>\n","protected":false},"template":"","tags":[],"research_cat":[2055],"class_list":["post-17856","research","type-research","status-publish","hentry","research_cat-faculty-of-information-technology"],"acf":{"research_manager":4999,"related_team":[17869,17870,17871,17747,17872,6668,17873,17875,17874]},"yoast_head":"\n