Sufficient Exploration for Convex Q-learning.
Fan LuPrashant G. MehtaSean MeynGergely NeuPublished in: CoRR (2022)
Keyphrases
- action selection
- exploration strategy
- cooperative
- reinforcement learning
- function approximation
- learning algorithm
- state space
- convex optimization
- model free
- multi agent
- convex hull
- stochastic approximation
- optimal policy
- temporal difference learning
- multi agent reinforcement learning
- strictly convex
- reinforcement learning algorithms
- globally optimal
- convex sets
- data sets
- convex relaxation
- state action
- piecewise linear
- dynamic environments
- risk minimization
- graph cuts
- dynamic programming
- convex constraints
- objective function
- active exploration