Domain Adaptation for Document Classification by Alternately Using Semi-supervised Learning and Feature Weighted Learning.
Hiroyuki ShinnouKanako KomiyaMinoru SasakiPublished in: PACLING (2017)
Keyphrases
- domain adaptation
- semi supervised learning
- document classification
- covariate shift
- labeled data
- semi supervised
- unlabeled data
- supervised learning
- co training
- transfer learning
- text categorization
- labeled and unlabeled data
- cross domain
- text classification
- weakly labeled data
- multiple sources
- active learning
- unsupervised learning
- sentiment classification
- learning process
- target domain
- opinion detection
- web documents
- learning tasks
- model selection
- machine learning
- test data
- training data
- data sets
- learning algorithm
- feature selection
- decision trees
- text documents
- data points
- small number
- text mining
- generative model