Loading...
Please wait, while we are loading the content...
Algorithms for Weighted Non-Negative Matrix Factorization
| Content Provider | CiteSeerX |
|---|---|
| Author | Blondel, Vincent Ho, Ngoc-Diep Dooren, Paul Van |
| Abstract | In this paper we introduce a new type of weighted non-negative matrix factorization and we show that the popular algorithms of Lee and Seung can easily be adapted to also incorporate such a weighting. We then prove that for appropriately chosen weighting matrices, the weighted Euclidean distance function and the weighted Kullback-Leibler divergence function are essentially identical. We finally show that the weighting can be chosen to emphasize parts of the data matrix to be approximated and this is applied successfully to the low rank fitting of a face image database. Keywords Non-negative matrix factorization, weighting, Euclidean distance, Kullback-Leibler divergence 1 |
| File Format | |
| Access Restriction | Open |
| Subject Keyword | Weighted Non-negative Matrix Factorization Weighted Euclidean Distance Function Euclidean Distance Low Rank Fitting Weighting Matrix Popular Algorithm Keywords Non-negative Matrix Factorization Data Matrix Weighted Kullback-leibler Divergence Function Kullback-leibler Divergence Face Image Database |
| Content Type | Text |
| Resource Type | Article |
