Semi-supervised NMF Models for Topic Modeling in Learning Tasks
This work addresses topic modeling for classification tasks, but it appears incremental as it builds on existing NMF methods with semi-supervised extensions.
The authors tackled the problem of topic modeling in learning tasks by proposing new semi-supervised nonnegative matrix factorization models, achieving high classification accuracy on the 20 Newsgroups dataset.
We propose several new models for semi-supervised nonnegative matrix factorization (SSNMF) and provide motivation for SSNMF models as maximum likelihood estimators given specific distributions of uncertainty. We present multiplicative updates training methods for each new model, and demonstrate the application of these models to classification, although they are flexible to other supervised learning tasks. We illustrate the promise of these models and training methods on both synthetic and real data, and achieve high classification accuracy on the 20 Newsgroups dataset.