Research map / Learning theory and algorithms / Combinatorial and search optimization
Continuous-Time Birth-Death MCMC for Bayesian Regression Tree Models
Reza Mohammadi, Matthew Pratola, Maurits Kaptein
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| add continuous-time birth-death sampling in place of discrete-time RJ-MCMC (ablation) | RJ-A · Synthetic regression tree (sigma^2=1) | effective sample size x2: 2370 -> 32374 | Improved |
| add rotation proposals to continuous-time sampler (ablation) | CT-A · Synthetic regression tree (sigma^2=1) | effective sample size x2: 32374 -> 74608 | Improved |
| add rotation proposals to discrete-time RJ-MCMC (ablation) | RJ-A · Synthetic regression tree (sigma^2=1) | effective sample size x2: 2370 -> 2899 | Improved |
| add perturbation of split rules to RJ-MCMC with rotation (ablation) | RJ-B · Synthetic regression tree (sigma^2=1) | effective sample size x1: 1419 -> 13134 | Improved |
| add perturbation proposals to continuous-time sampler with rotation (ablation) | CT-B · Synthetic regression tree (sigma^2=1) | unique trees visited: 3.82 -> 11.71 | Improved |
| use continuous-time instead of discrete-time sampler, both with rotation and perturbation (ablation) | RJ-C · Synthetic regression tree (sigma^2=1) | unique trees visited: 8.29 -> 11.71 | Improved |
| use continuous-time sampler instead of RJ-MCMC | RJ-A · Synthetic regression tree (sigma^2=1) | test MSE: 1.02 -> 1.02 | No clear effect |
| add continuous-time birth-death sampling in place of discrete-time RJ-MCMC (ablation) | RJ-A · Synthetic regression tree (sigma^2=1) | effective samples per second x2: 3160 -> 9018 | Improved |
| use continuous-time sampler instead of RJ-MCMC | RJ-A · Synthetic regression tree (sigma^2=1) | effective samples per second sigma^2: 26344 -> 5371 | Hurt |
| use continuous-time sampler in lower-noise setting sigma^2=0.01 | RJ-A · Synthetic regression tree (sigma^2=0.01) | effective sample size x2: 3029 -> 21063 | Improved |
Closest papers by meaning
- Monte-Carlo Tree Search with Uncertainty Propagation via Optimal Transport (ICML 2025 · similarity 0.85)
- Improving Monte Carlo Tree Search for Symbolic Regression (NeurIPS 2025 · similarity 0.84)
- Optimal Survival Trees: A Dynamic Programming Approach (AAAI 2024 · similarity 0.84)
- Power Mean Estimation in Stochastic Continuous Monte-Carlo Tree Search (ICML 2025 · similarity 0.83)
- BARK: A Fully Bayesian Tree Kernel for Black-box Optimization (ICML 2025 · similarity 0.83)
- Monte Carlo Tree Search in the Presence of Transition Uncertainty (AAAI 2024 · similarity 0.83)
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.
