Can Nonnegative Matrix Factorization be used for extracting the most important features from a dataset with high dimensionality (1000+ variables)?


I would like to extract and identify the most important variables from a biological dataset containing mutational and gene expression data in R. Previous supervised techniques showed that these features have weak predictability. Can I use NNMF for this aim?


Posted 2020-12-08T19:54:48.757

Reputation: 1

No answers