A Novel Graph-Based Estimation of the Distribution Algorithm and its Extension Using Reinforcement Learning.
Xianneng LiShingo MabuKotaro HirasawaPublished in: IEEE Trans. Evol. Comput. (2014)
Keyphrases
- learning algorithm
- reinforcement learning
- estimation algorithm
- experimental evaluation
- objective function
- times faster
- detection algorithm
- np hard
- high accuracy
- worst case
- probabilistic model
- dynamic programming
- computational complexity
- optimal solution
- significant improvement
- probability distribution
- function approximation
- optimization algorithm
- estimation accuracy
- cost function
- density function
- convergence rate
- matching algorithm
- optimal policy
- preprocessing
- expectation maximization
- search space
- k means
- state space
- stochastic approximation
- graph based algorithm
- kernel density
- model free
- clustering method
- segmentation algorithm
- em algorithm
- computational cost