ИСТИНА

Invited talk (30 minutes) is devoted to statistical problems of identification of the most significant combinations of genetic and nongenetic factors which could increase the risk of a complex disease. New theorems are proved to justify different methods of data analysis in the framework of specified stochastic models. The main attention will be paid to the new version of the MDR-method, various modifications of logic regression and machine learning methods. We apply the techniques of random trees, stochastic optimization and the random fields theory. The applications to analysis of cardio-vascular diseases are provided as well.