Robust, Efficient, Globally-Optimized Reinforcement Learning with the Parti-Game Algorithm.
Mohammad A. Al-AnsariRonald J. WilliamsPublished in: NIPS (1998)
Keyphrases
- computationally efficient
- single pass
- dynamic programming
- optimization algorithm
- experimental evaluation
- reinforcement learning
- high accuracy
- learning algorithm
- similarity measure
- high efficiency
- k means
- worst case
- convergence rate
- parameter tuning
- memory efficient
- matching algorithm
- expectation maximization
- genetic algorithm
- significant improvement
- objective function
- segmentation algorithm
- detection algorithm
- np hard
- highly efficient
- computational complexity
- data structure
- pruning strategy
- control policy