SACHA: Soft Actor-Critic With Heuristic-Based Attention for Partially Observable Multi-Agent Path Finding.
Qiushi LinHang MaPublished in: IEEE Robotics Autom. Lett. (2023)
Keyphrases
- partially observable
- path finding
- reinforcement learning
- multi agent
- single agent
- partially observable markov decision processes
- state space
- policy gradient
- dynamical systems
- heuristic search
- markov decision processes
- decision problems
- path planning
- function approximation
- temporal difference
- reinforcement learning algorithms
- policy iteration
- infinite horizon
- average reward
- search algorithm
- optimal path
- belief state
- dynamic programming
- hill climbing
- rule learning
- multiple agents
- model free
- learning algorithm
- reward function
- markov chain
- mobile robot
- finite state
- optimal policy
- computational complexity
- multi agent systems
- machine learning