Semi-supervised Nonnegative Matrix Factorization for Document Classification
Semi-supervised Nonnegative Matrix Factorization for Document Classification
复制标题
DOI：
10.1109/ieeeconf53345.2021.9723109
复制
发表时间：
2021-10
期刊：
影响因子：
--
通讯作者：
Jamie Haddock;Lara Kassab;Sixian Li;Alona Kryshchenko;Rachel Grotheer;Elena Sizikova;Chuntian Wang;Thomas Merkh;R. W. M. A. Madushani;Miju Ahn;D. Needell;Kathryn Leonard
中科院分区：
文献类型：
--
作者：
Jamie Haddock;Lara Kassab;Sixian Li;Alona Kryshchenko;Rachel Grotheer;Elena Sizikova;Chuntian Wang;Thomas Merkh;R. W. M. A. Madushani;Miju Ahn;D. Needell;Kathryn Leonard
我们提出了用于文档分类的新半监督非负矩阵分解（SSNMF）模型，并为这些模型作为最大似然估计器提供了动机。所提出的 SSNMF 模型同时提供主题模型和分类模型，从而提供高度可解释的分类结果。我们对每个新模型使用乘法更新导出训练方法，并演示这些模型在单标签和多标签文档分类中的应用，尽管这些模型对于其他监督学习任务（例如回归）很灵活。我们在文档分类数据集（例如 20 个新闻组、路透社）上阐述了这些模型和训练方法的前景。
We propose new semi-supervised nonnegative matrix factorization (SSNMF) models for document classification and provide motivation for these models as maximum likelihood estimators. The proposed SSNMF models simultaneously provide both a topic model and a model for classification, thereby offering highly interpretable classification results. We derive training methods using multiplicative updates for each new model, and demonstrate the application of these models to single-label and multi-label document classification, although the models are flexible to other supervised learning tasks such as regression. We illustrate the promise of these models and training methods on document classification datasets (e.g., 20 Newsgroups, Reuters).