Optimal Sample Complexity of Contrastive Learning.
Noga AlonDmitrii AvdiukhinDor ElboimOrr FischerGrigory YaroslavtsevPublished in: ICLR (2024)
Keyphrases
- sample complexity
- learning problems
- learning algorithm
- supervised learning
- dynamic programming
- decision lists
- active learning
- learning tasks
- linear threshold
- prior knowledge
- pac learning
- learning process
- upper bound
- training data
- theoretical analysis
- data sets
- pac learnability
- pac model
- small number
- special case
- feature space
- reinforcement learning
- machine learning
- data mining