You Only Evaluate Once: a Simple Baseline Algorithm for Offline RL.
Wonjoon GooScott NiekumPublished in: CoRL (2021)
Keyphrases
- times faster
- computational complexity
- learning algorithm
- experimental evaluation
- significant improvement
- k means
- preprocessing
- search space
- optimal solution
- cost function
- dynamic programming
- worst case
- monte carlo
- detection algorithm
- improved algorithm
- convergence rate
- model free
- optimization algorithm
- segmentation algorithm
- computationally efficient
- high accuracy
- probabilistic model
- similarity measure
- neural network