Stronger Generalization Bounds for Deep Nets via a Compression Approach.
Sanjeev AroraRong GeBehnam NeyshaburYi ZhangPublished in: ICML (2018)
Keyphrases
- generalization bounds
- data dependent
- learning theory
- generalization ability
- model selection
- convex combinations
- ranking algorithm
- vc dimension
- linear classifiers
- compression scheme
- learning problems
- data compression
- statistical learning theory
- compression ratio
- compression algorithm
- image compression
- kernel machines
- learning machines
- ranking functions
- lower bound
- pairwise
- ensemble learning
- statistical learning
- support vector machine svm
- ensemble methods