Self-training Converts Weak Learners to Strong Learners in Mixture Models.
Spencer FreiDifan ZouZixiang ChenQuanquan GuPublished in: AISTATS (2022)
Keyphrases
- mixture model
- weak learners
- em algorithm
- gaussian mixture model
- ensemble learning
- model selection
- generative model
- generalization error
- unsupervised learning
- probabilistic model
- learning process
- expectation maximization
- multi class classification
- language model
- boosting algorithms
- learning environment
- cost sensitive
- semi supervised
- semi supervised learning
- linear classifiers
- training set
- weak classifiers
- maximum likelihood
- pattern recognition