Reward Conditioned Neural Movement Primitives for Population-Based Variational Policy Optimization.
M. Tuluhan AkbulutUtku BozdoganAhmet E. TekdenEmre UgurPublished in: ICRA (2021)
Keyphrases
- combinatorial optimization
- optimization algorithm
- optimal policy
- reward function
- reinforcement learning
- optimization problems
- average reward
- network architecture
- optimization process
- neural network
- global optimization
- differential evolution
- partially observable environments
- neural model
- optimization method
- particle swarm optimization
- high level
- simulated annealing
- supply chain
- state space
- optical flow
- bio inspired
- artificial neural networks
- multiscale
- evolutionary search
- policy gradient
- inverse reinforcement learning
- image segmentation