Learning without the Phase: Regularized PhaseMax Achieves Optimal Sample Complexity.
Fariborz SalehiEhsan AbbasiBabak HassibiPublished in: NeurIPS (2018)
Keyphrases
- sample complexity
- learning problems
- decision lists
- learning algorithm
- linear threshold
- active learning
- training data
- reinforcement learning
- supervised learning
- learning tasks
- unsupervised learning
- concept learning
- dynamic programming
- learning process
- concept classes
- optimal solution
- objective function
- pac learnability
- sample complexity bounds