MVICAD2: Multi-View Independent Component Analysis with Delays and Dilations
· By Antonio Sedino, CTRO · Published by Reinventy Solutions Corp.
Machine learning techniques in multi-view settings face challenges integrating heterogeneous data, aligning feature spaces, and managing view-specific biases, particularly in neuroscience applications.
Apple Machine Learning Research introduces MVICAD2, a method for multi-view independent component analysis incorporating delays and dilations. The approach addresses challenges in aligning heterogeneous data and managing view-specific biases, especially in neuroscience group studies using MEG data.
Read the original source at Apple Machine Learning Research ↗