Supervised Quantile Normalization for Matrix Factorization Using Optimal Transport
Save Big on Coursera Plus. 7,000+ courses at $160 off. Limited Time Only! Grab it Explore a seminar on supervised quantile normalization for matrix factorization using optimal transport. Delve into the speaker's recent work applying regularized optimal transport to perform "soft" sorting and ranking. Learn about the expansion of this framework to include differentiable "soft" quantile normalization operators and their application to dimensionality reduction. Discover algorithms for normalizing features with target quantile distributions to obtain matrices that are easier to factorize. Examine empirical evidence of recovery and practical applications in genomics. Gain insights from Marco Cuturi, a Google Brain researcher with extensive experience in applied mathematics, statistics, and machine learning, particularly in optimal transport and Wasserstein distances.