A New Lower Bound for Agnostic Learning with Sample Compression Schemes.
Steve HannekeAryeh KontorovichPublished in: CoRR (2018)
Keyphrases
- compression scheme
- agnostic learning
- concept class
- lower bound
- vc dimension
- image compression
- concept classes
- upper bound
- compression ratio
- uniform distribution
- data compression
- membership queries
- target function
- learning theory
- sample size
- compression algorithm
- pac learning
- noise tolerant
- objective function
- optimal solution
- dnf formulas
- image processing
- lower and upper bounds
- complexity analysis
- pac model
- decision lists
- upper and lower bounds
- sample complexity
- boosting algorithms
- random sampling
- efficient learning
- concept learning
- boolean functions
- wavelet transform
- support vector
- similarity measure
- computer vision