Patching Approximate Solutions in Reinforcement Learning.
Min Sub KimWilliam T. B. UtherPublished in: ECML (2006)
Keyphrases
- approximate solutions
- reinforcement learning
- np hard
- function approximation
- state space
- optimal solution
- hard optimization problems
- reinforcement learning algorithms
- exact solution
- markov decision processes
- model free
- temporal difference
- supervised learning
- energy function
- optimal policy
- learning algorithm
- dynamic programming
- computer vision
- temporal difference learning
- reinforcement learning methods
- policy search
- machine learning
- graph cuts
- genetic programming
- learning classifier systems
- simulated annealing
- learning process
- multi agent
- partially observable markov decision processes