Feature Selection through Minimization of the VC dimension.
JayadevaSanjit S. BatraSiddarth SabharwalPublished in: CoRR (2014)
Keyphrases
- vc dimension
- feature selection
- sample complexity
- upper bound
- covering numbers
- sample size
- lower bound
- statistical learning theory
- concept classes
- distribution free
- vapnik chervonenkis dimension
- generalization bounds
- worst case
- text categorization
- pac learning
- concept class
- objective function
- inductive inference
- uniform convergence
- support vector
- euclidean space
- support vector machine
- feature space
- theoretical analysis
- compression scheme
- text classification
- dimensionality reduction
- multi class
- learning algorithm
- machine learning
- unsupervised learning
- special case
- feature extraction