Winner Takes It All: Training Performant RL Populations for Combinatorial Optimization.
Nathan GrinsztajnDaniel Furelos-BlancoShikha SuranaClément BonnetTom BarrettPublished in: NeurIPS (2023)
Keyphrases
- combinatorial optimization
- traveling salesman problem
- metaheuristic
- combinatorial optimization problems
- simulated annealing
- reinforcement learning
- combinatorial problems
- mathematical programming
- branch and bound algorithm
- optimization problems
- branch and bound
- hard combinatorial optimization problems
- state space
- training set
- vehicle routing problem
- path relinking
- tabu search
- optimal policy
- dynamic programming
- multi objective
- function approximation
- multiscale
- memetic algorithm
- image segmentation
- exact algorithms
- combinatorial search
- submodular functions
- neural network