Solving Sokoban with Forward-Backward Reinforcement Learning.
Yaron ShohamGal ElidanPublished in: SOCS (2021)
Keyphrases
- forward backward
- reinforcement learning
- hidden markov models
- state space
- problems in artificial intelligence
- function approximation
- optimal policy
- combinatorial optimization
- markov decision processes
- control problems
- quadratic programming
- reinforcement learning algorithms
- learning algorithm
- machine learning
- neural network
- data sets
- mobile robot
- optimal control
- model free
- lower bound
- multi agent
- partially observable
- reinforcement learning agents