Training Reinforcement Neurocontrollers Using the Polytope Algorithm.
Aristidis LikasIsaac E. LagarisPublished in: Neural Process. Lett. (1999)
Keyphrases
- dynamic programming
- experimental evaluation
- learning algorithm
- detection algorithm
- recognition algorithm
- preprocessing
- k means
- significant improvement
- high accuracy
- training algorithm
- search space
- computational cost
- convergence rate
- times faster
- theoretical analysis
- expectation maximization
- simulated annealing
- cost function
- computational complexity
- matching algorithm
- active learning
- knapsack problem
- optimal solution
- reinforcement learning