New paper proposes Bayesian transfer for offline Q-learning using multi-task Gaussian processes
A new arXiv paper presents a Bayesian transfer approach for offline reinforcement learning that uses multi-task Gaussian processes to leverage data from related tasks while guarding against negative transfer.
Read the full story at arXiv stat.ML ↗
Timeline · 1 report
- 2026-11-10 05:00 · arXiv stat.ML
Bayesian transfer for offline Q-learning with multi-task Gaussian processes