L1 vs. L2 Regularization in Text Classification when Learning from Labeled Features.
Sinziana MaziluJosé IriaPublished in: ICMLA (1) (2011)
Keyphrases
- text classification
- learning algorithm
- learning process
- supervised learning
- online learning
- positive and unlabeled examples
- feature engineering
- learning tasks
- feature vectors
- active learning
- unsupervised learning
- bag of words
- text categorization
- neural network
- reinforcement learning
- feature extraction
- training examples
- action recognition
- prior knowledge
- learning problems
- multi task
- learned models
- labeled training data
- information retrieval