Modified Q-learning method with fuzzy state division and adaptive rewards.
Yoichiro MaedaPublished in: FUZZ-IEEE (2002)
Keyphrases
- dynamic programming
- high accuracy
- preprocessing
- detection method
- reinforcement learning
- pairwise
- evaluation method
- learning algorithm
- fuzzy clustering
- fuzzy membership functions
- fuzzy reasoning
- fuzzy rules
- support vector machine svm
- denoising
- probabilistic model
- significant improvement
- feature space
- computational complexity