Collision Cross-entropy and EM Algorithm for Self-labeled Classification.
Zhongwen ZhangYuri BoykovPublished in: CoRR (2023)
Keyphrases
- em algorithm
- cross entropy
- log likelihood
- maximum likelihood
- expectation maximization
- unlabeled samples
- mixture model
- density estimation
- maximum likelihood estimation
- gaussian mixture model
- parameter estimation
- generative model
- supervised learning
- training set
- gaussian mixture
- classification accuracy
- machine learning
- maximum a posteriori
- hidden variables
- unsupervised learning
- semi supervised
- probability density function
- support vector machine
- training data
- text classification
- support vector
- decision trees
- feature selection
- class labels
- feature vectors
- feature space
- scoring function
- image segmentation