A kernel-based model-free RL method combines Bernstein bonuses with non-parametric smoothing to improve the horizon dependence in finite-horizon regret bounds.
PhD thesis, Carnegie Mellon University (2024)
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
citation-role summary
background 1
citation-polarity summary
fields
cs.LG 1years
2026 1verdicts
UNVERDICTED 1roles
background 1polarities
background 1representative citing papers
citing papers explorer
-
Improved Model-based Reinforcement Learning with Smooth Kernels
A kernel-based model-free RL method combines Bernstein bonuses with non-parametric smoothing to improve the horizon dependence in finite-horizon regret bounds.