Research map / Learning theory and algorithms / Combinatorial and search optimization
Beyond Expected Values: Risk-Sensitive Planning with Distributional Monte-Carlo Tree Search
Read the paperCheck your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's abstract only. Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Proposes CATSO and PATSO, distributional MCTS variants that maintain finite-support empirical backup distributions at Q-edges to support risk-sensitive objectives such as CVaR.
Problem
Standard MCTS stores only scalar mean value estimates, making it poorly suited for tail-risk objectives such as conditional value-at-risk (CVaR).
Approach
Maintains finite-support empirical backup laws at Q-edges within MCTS, using categorical (CATSO) or particle (PATSO) representations combined with Thompson sampling with optimism, so that tree search can optimize distributional, risk-sensitive objectives rather than expected values.
Experimental setup
Builds on: Monte-Carlo Tree Search Thompson sampling
Closest papers by meaning
- Monte Carlo Tree Search in the Presence of Transition Uncertainty (AAAI 2024 · similarity 0.88)
- Threshold UCT: Cost-Constrained Monte Carlo Tree Search with Pareto Curves (AAAI 2025 · similarity 0.88)
- Graph-Based Stochastic-Power-UCT: Monte-Carlo Graph Search with Power Mean Estimation (NeurIPS 2026 · similarity 0.87)
- A Bayesian Approach to Online Planning (ICML 2024 · similarity 0.87)
- Monte-Carlo Tree Search with Uncertainty Propagation via Optimal Transport (ICML 2025 · similarity 0.87)
- Online Robust Reinforcement Learning Through Monte-Carlo Planning (ICML 2025 · similarity 0.87)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.