Использование гауссовских процессов для выявления связи между генотипом и фенотипом в данных Enroll-HD

In this work we consider Functional Dimensionality Reduction for Gaussian Processes applied for investigation of Huntington Disease. We are going to give short theoretical overview of the method, remind what is Gaussian Processes, discuss PCA and Sparse PCA, consider methods of calculation explained variance in case of not orthogonal loads. Also we will look at synthetic samples of using EDR-GP and present our library edr-gp. The major part of this work is dedicated to results of applying this approach to EnrollHD database.

Результаты своих текущих проектов представили сотрудники сектора "Анализ данных в нейронауках" Артем Борзов и Николай Луцяк.