Graph-based Deterministic Policy Gradient for Repetitive Combinatorial Optimization Problems.
Zhongyuan ZhaoAnanthram SwamiSantiago SegarraPublished in: ICLR (2023)
Keyphrases
- combinatorial optimization problems
- policy gradient
- optimization problems
- combinatorial optimization
- knapsack problem
- ant colony optimization
- metaheuristic
- reinforcement learning
- function approximation
- optimal control
- traveling salesman problem
- gradient method
- semi supervised
- reinforcement learning algorithms
- approximation methods
- vehicle routing problem
- average reward
- variance reduction
- lower bound
- computational complexity
- partially observable markov decision processes
- reinforcement learning methods
- neural network