A Meta-Preconditioning Approach for Deep Q-Learning.
Spilios EvmorfosAthina P. PetropuluPublished in: ICASSP (2024)
Keyphrases
- reinforcement learning
- function approximation
- cooperative
- edge preserving
- multi agent
- state space
- learning algorithm
- optical flow computation
- meta level
- bundle adjustment
- multi agent reinforcement learning
- stochastic approximation
- model free
- learning rate
- potential field
- deep learning
- reinforcement learning algorithms
- neural network
- conjugate gradient
- optimal policy
- primal dual
- action selection
- iterative methods
- least squares
- structure from motion
- optical flow
- temporal difference learning
- meta reasoning