Reinforcement learning control based on multi-goal representation using hierarchical heuristic dynamic programming.
Zhen NiHaibo HeDongbin ZhaoDanil V. ProkhorovPublished in: IJCNN (2012)
Keyphrases
- dynamic programming
- reinforcement learning
- optimal control
- control problems
- state space
- coarse to fine
- optimal policy
- hierarchical reinforcement learning
- control policy
- linear programming
- markov decision processes
- hierarchical representation
- policy search
- stereo matching
- control system
- dynamic programming algorithms
- robot control
- knapsack problem
- action selection
- optimal solution
- lagrangian relaxation
- temporal difference
- infinite horizon
- control method
- function approximation
- single machine
- greedy algorithm
- hierarchical structure
- search algorithm
- adaptive control
- control strategy
- machine learning
- multi agent
- greedy search
- inverted pendulum
- image representation
- search space
- learning process