Generalized Hierarchical Bayesian Segmentation with Irregular Designs, Multi-Sequence Hierarchies, and Grouped/Latent-Group Designs

arXiv:2603.14681 · cs.LG · Submitted 2026-03-16 · Read on arXiv

cs.LG

Submitted: 2026-03-16

Updated: 2026-10-28

DOI: 10.5281/zenodo.23711344

Code: https://github.com/osolari/bayesbreak

Project page: https://cloud.r-project.org

License: http://creativecommons.org/licenses/by/4.0/

The gist: Bayesian change-point and segmentation models provide uncertainty-aware piecewise-constant representations of ordered data, but exact inference is often limited to narrow likelihood classes…

Terminology

Abstract

Bayesian change-point and segmentation models provide uncertainty-aware piecewise-constant representations of ordered data, but exact inference is often limited to narrow likelihood classes, single sequences, or index-uniform designs. We present BayesBreak, a modular offline Bayesian segmentation framework that separates local block scoring from global inference: each candidate block supplies a marginal likelihood and any needed moment numerators, while a dynamic program combines these scores to compute posteriors over segment counts, boundaries, and latent signals. For weighted exponential-family likelihoods with conjugate priors, block evidences and posterior moments are available in closed form from cumulative sufficient statistics, enabling exact sum-product inference for p(yk), p(ky), boundary marginals, and Bayes regression curves. We distinguish these summaries from the joint MAP segmentation, recovered by a separate max-sum recursion. BayesBreak supports design-aware partition priors for irregular observations, exact pooling across replicates with shared boundaries, and latent-template mixtures with exact EM updates. For non-conjugate GLM blocks, the same DP layer can use deterministic local approximations such as Laplace, variational methods, EP, or quadrature. We prove a posterior-odds stability bound: uniform per-block log-evidence error ε perturbs k-odds and boundary-odds by at most (k+k')ε and 2kε. Validation includes synthetic recovery, calibration, and scaling experiments, plus four real-data illustrations: well-log geology, array-CGH copy number, equity-return volatility, and CpG-atlas methylation.

Sources

Related papers