%0 Generic %1 nguyen2016parallel %A Nguyen, Duy Khuong %A Ho, Tu Bao %D 2016 %K NNMF acreuser %T Fast Parallel Randomized Algorithm for Nonnegative Matrix Factorization with KL Divergence for Large Sparse Datasets %U http://arxiv.org/abs/1604.04026 %X Nonnegative Matrix Factorization (NMF) with Kullback-Leibler Divergence (NMF-KL) is one of the most significant NMF problems and equivalent to Probabilistic Latent Semantic Indexing (PLSI), which has been successfully applied in many applications. For sparse count data, a Poisson distribution and KL divergence provide sparse models and sparse representation, which describe the random variation better than a normal distribution and Frobenius norm. Specially, sparse models provide more concise understanding of the appearance of attributes over latent components, while sparse representation provides concise interpretability of the contribution of latent components over instances. However, minimizing NMF with KL divergence is much more difficult than minimizing NMF with Frobenius norm; and sparse models, sparse representation and fast algorithms for large sparse datasets are still challenges for NMF with KL divergence. In this paper, we propose a fast parallel randomized coordinate descent algorithm having fast convergence for large sparse datasets to archive sparse models and sparse representation. The proposed algorithm's experimental results overperform the current studies' ones in this problem.

@misc{nguyen2016parallel, abstract = {Nonnegative Matrix Factorization (NMF) with Kullback-Leibler Divergence (NMF-KL) is one of the most significant NMF problems and equivalent to Probabilistic Latent Semantic Indexing (PLSI), which has been successfully applied in many applications. For sparse count data, a Poisson distribution and KL divergence provide sparse models and sparse representation, which describe the random variation better than a normal distribution and Frobenius norm. Specially, sparse models provide more concise understanding of the appearance of attributes over latent components, while sparse representation provides concise interpretability of the contribution of latent components over instances. However, minimizing NMF with KL divergence is much more difficult than minimizing NMF with Frobenius norm; and sparse models, sparse representation and fast algorithms for large sparse datasets are still challenges for NMF with KL divergence. In this paper, we propose a fast parallel randomized coordinate descent algorithm having fast convergence for large sparse datasets to archive sparse models and sparse representation. The proposed algorithm's experimental results overperform the current studies' ones in this problem.}, added-at = {2016-04-15T06:39:01.000+0200}, author = {Nguyen, Duy Khuong and Ho, Tu Bao}, biburl = {https://www.bibsonomy.org/bibtex/2ba294538cb4245b7e4d190b937cd6dca/pixor}, description = {1604.04026v1.pdf}, interhash = {026c87e8a9702011c1706c4f7aaad1b0}, intrahash = {ba294538cb4245b7e4d190b937cd6dca}, keywords = {NNMF acreuser}, note = {cite arxiv:1604.04026v1.pdf}, timestamp = {2016-04-15T06:39:01.000+0200}, title = {Fast Parallel Randomized Algorithm for Nonnegative Matrix Factorization with KL Divergence for Large Sparse Datasets}, url = {http://arxiv.org/abs/1604.04026}, year = 2016 }