Learners that Use Little Information

Keyphrases

• Learning Algorithm 100%
• Empirical Risk Minimization 100%
• Tight 50%
• Learnability 50%
• Error Parameter 50%
• Sample Complexity 50%
• Simple Concept 50%
• Differential Privacy 50%
• Concept Classes 50%
• Complexity Bounds 50%
• Occam's Razor 50%
• PAC-Bayes 50%
• Sample Compression Schemes 50%
• Input Samples 50%
• VC-classes 50%
• Confidence Parameters 50%
• Distribution Dependence 50%

Computer Science

• Learning Algorithm 100%
• Compression Scheme 50%
• Differential Privacy 50%
• Information Bit 50%
• Learning Capacity 50%

Mathematics

• Upper Bound 100%
• Occam's Razor 100%