Tree Bandits for Generative Bayes: Simulation Outputs for ABC and MAP Estimation

Sean O’Hagan's research outputs from a 2026 paper proposing a self-aware framework to accelerate Approximate Bayesian Computation (ABC) rejection sampling. The dataset includes code, supplementary materials, and likely contains simulation results from applications in epidemiological modeling and alloy material discovery. The 4.6 MB collection is licensed under CC-BY-4.0.

Use Cases

• Posterior sampling in epidemiological models based on the described ABC-Tree method
• Model-based Bayesian optimization for material discovery based on the described MAP-Tree method
• Benchmarking new ABC acceleration algorithms against the provided simulation outputs
• Studying the application of recursive partitioning and bandit algorithms to likelihood-free inference

Strengths

• Includes source code and supplementary files for reproducibility
• Licensed under the permissive CC-BY-4.0 license
• File formats suggest a mix of data (CSV, NPZ), code (PY, IPYNB), and documentation (PDF, MD)

Limitations

• Row count and specific column definitions are unknown, limiting suitability assessment
• The dataset's primary content appears to be code and experiment outputs rather than a standalone data corpus
• Description metadata is limited; actual data quality requires manual inspection after download

Provenance

Source

figshare

Collection Method

Generated as part of academic research on accelerating Approximate Bayesian Computation.

Time Range

null

Freshness

Last updated 2026-05-01 14:35:02

Geography

null