Backtracking Restarts for Deep Reinforcement Learning.
Zaid MarjiJohn LicatoPublished in: FLAIRS Conference (2021)
Keyphrases
- reinforcement learning
- backtracking search
- clause learning
- function approximation
- search algorithm
- constraint satisfaction
- random walk
- reinforcement learning algorithms
- model free
- search space
- state space
- multi agent reinforcement learning
- temporal difference
- optimal policy
- constraint satisfaction problems
- dependency directed backtracking
- constraint propagation
- deep learning
- search problems
- markov decision processes
- backtracking algorithm
- multi agent
- dynamic programming
- temporal difference learning
- stochastic approximation
- transition model
- reinforcement learning methods
- optimal control
- markov decision process
- partially observable
- reward function
- data sets
- search tree
- satisfiability problem
- supervised learning
- least squares
- active learning
- learning algorithm