Deep Reinforcement Learning for Smart Restarts in Exploration-Only Exploitation-Only Hybrid Metaheuristics.
Antonio Bolufé RöhlerBowen XuPublished in: MIC (2) (2024)
Keyphrases
- reinforcement learning
- exploration exploitation tradeoff
- active exploration
- function approximation
- action selection
- metaheuristic
- autonomous learning
- relevance feedback
- model free
- exploration exploitation
- state space
- random walk
- exploration strategy
- particle swarm optimization
- model based reinforcement learning
- machine learning
- objective function
- optimization problems
- ant colony optimization
- search capabilities
- temporal difference
- grasp with path relinking
- combinatorial optimization
- hybrid learning
- transfer learning
- heuristic search
- active learning
- learning algorithm
- smart environments
- learning tasks
- markov decision processes
- action space
- stochastic approximation
- dynamic programming
- evolutionary algorithm
- multi agent