Algorithms, Initializations, and Convergence for the Nonnegative Matrix Factorization
arXiv:1407.7299
Abstract
It is well known that good initializations can improve the speed and accuracy of the solutions of many nonnegative matrix factorization (NMF) algorithms. Many NMF algorithms are sensitive with respect to the initialization of W or H or both. This is especially true of algorithms of the alternating least squares (ALS) type, including the two new ALS algorithms that we present in this paper. We compare the results of six initialization procedures (two standard and four new) on our ALS algorithms. Lastly, we discuss the practical issue of choosing an appropriate convergence criterion.
References in corpus (1)
Cited by in corpus (16)
- Data Fusion by Matrix Factorization
- A Field Guide to Forward-Backward Splitting with a FASTA Implementation
- NIMFA: A Python Library for Nonnegative Matrix Factorization
- Randomized Nonnegative Matrix Factorization
- A Case Study in Text Mining: Interpreting Twitter Data From World Cup Tweets
- Initialization for Nonnegative Matrix Factorization: a Comprehensive Review
- n-stage Latent Dirichlet Allocation: A Novel Approach for LDA
- Pattern Decomposition with Complex Combinatorial Constraints: Application to Materials Discovery
- Algorithms for Internal Validation Clustering Measures in the Post Genomic Era
- Semi-orthogonal Non-negative Matrix Factorization with an Application in Text Mining
- Clustering and Latent Semantic Indexing Aspects of the Nonnegative Matrix Factorization
- Privacy-Preserving Multiple Tensor Factorization for Synthesizing Large-Scale Location Traces with Cluster-Specific Features
- HYPERION: Hyperspectral Penetrating-type Ellipsoidal Reconstruction for Terahertz Blind Source Separation
- Prediction of Cancer Microarray and DNA Methylation Data using Non-negative Matrix Factorization
- Supervised Nonnegative Matrix Factorization to Predict ICU Mortality Risk
- Extending the Unmixing methods to Multispectral Images