Incremental Feature Selection and l1 Regularization for Relaxed Maximum-Entropy Modeling.
Stefan RiezlerAlexander VassermanPublished in: EMNLP (2004)
Keyphrases
- maximum entropy modeling
- feature selection
- maximum entropy models
- incremental learning
- text categorization
- support vector
- feature selection algorithms
- mutual information
- machine learning
- classification accuracy
- high dimensionality
- feature subset
- dimensionality reduction
- model selection
- support vector machine
- information gain
- method for feature selection
- feature ranking
- feature space
- selected features
- text classification
- regularization parameter
- microarray data
- irrelevant features
- unsupervised feature selection
- naive bayes
- multi class