A probabilistic matrix factorization method that prunes graph edges with negative latent-feature correlations improves prediction accuracy and keeps computation linear in the number of nonzeros.
Metagenes and molecular pattern discovery using matrix factorization
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.LG 1years
2019 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
Scalable Probabilistic Matrix Factorization with Graph-Based Priors
A probabilistic matrix factorization method that prunes graph edges with negative latent-feature correlations improves prediction accuracy and keeps computation linear in the number of nonzeros.