An Iterative Algorithm for Regularized Non-negative Matrix Factorizations
arXiv:2410.22698v1h-index: 8
Originality Synthesis-oriented
AI Analysis
This work provides an incremental improvement to matrix factorization methods for data analysis tasks.
The authors tackled the problem of improving non-negative matrix factorization by introducing a weighted norm and regularization, resulting in an additive update algorithm that avoids zero-value stagnation and was applied to cocktail database reduction.
We generalize the non-negative matrix factorization algorithm of Lee and Seung to accept a weighted norm, and to support ridge and Lasso regularization. We recast the Lee and Seung multiplicative update as an additive update which does not get stuck on zero values. We apply the companion R package rnnmf to the problem of finding a reduced rank representation of a database of cocktails.