arXiv subjects

Yuh-Haur Chen

Publications and source records attributed to Yuh-Haur Chen.

3 recordsLinked to original sources

arXiv2026

A Bayesian Adaptive Latent Mixture Model for Zero-Inflated Weighted Brain Connectome Analysis

Replicated weighted networks often exhibit many structural zeros alongside heterogeneous non-zero edge strengths. In structural connectomics, this zero-inflation coincides with subjects expressing overlapping, rather than discrete, connectivity patterns. To address these features, we propose a Bayesian adaptive latent mixture model for zero-inflated weighted networks. Our approach represents each subject network as a simplex mixture of shared low-rank latent score matrices, integrated with a hurdle likelihood that separates edge existence from conditional edge strength. A sparsity-coupling parameter enables absent edges to be either independent of, or informative about, the latent connectivity. For computation, we employ transformed Hamiltonian Monte Carlo on unconstrained coordinates, selecting the number of templates via predictive fit, held-out link prediction, and template stability. Theoretically, we establish posterior consistency, local asymptotic normality, a Bernstein--von Mises approximation, and predictive consistency for an identifiable quotient-space estimand under a fixed-template scenario. Simulations demonstrate performance gains over topology-only baselines in settings with mixed memberships or structure-informed sparsity. Applied to Human Connectome Project data, the model recovers stable latent score patterns and heterogeneous subject-level mixtures, with behavioural analyses serving strictly as exploratory annotations rather than confirmatory biomarker claims.

stat.ME↗

arXiv2026

Bayesian Deep Count Regression and Anomaly Detection: Evidence from GDELT Event Panels

The Global Database of Events, Language and Tone (GDELT) provides geolocated event records that can be aggregated into weekly spatiotemporal panels of event counts across regions, actors, and event types. These panels are typically sparse, bursty, and overdispersed, so calibrated probabilistic forecasting is essential for monitoring rare surges. We propose Bayesian count regression pipelines that pair deterministic deep temporal encoders with negative binomial (NB2) and zero-inflated negative binomial (ZINB2) likelihood heads. Posterior predictive simulation yields predictive quantiles and right-tail probabilities that support both forecasting and anomaly scoring. For interpretable spillover attribution, we also fit a Bayesian generalised linear model with high-dimensional lagged cross-series predictors and a two-step screen-and-refit procedure under a three-parameter beta-normal (TPBN) shrinkage prior. To connect spillovers to directional statistics, active cross-region effects are mapped to geodesic bearings on the World Geodetic System 1984 ellipsoid (WGS84) and summarised using weighted circular moments, rose diagrams, and bearing-field maps. Simulations with known spillovers and conflict-panel case studies show accurate right-tail behaviour and a practical workflow for detecting and interpreting geopolitical shocks.

stat.AP↗

arXiv2025

Cyclic-Shift Sparse Kronecker Tensor Classifier for Signal-Region Detection in Neuroimaging

This study proposes a cyclic-shift logistic sparse Kronecker product decomposition (SKPD) model for high-dimensional tensor data, enhancing the SKPD framework with a cyclic-shift mechanism for binary classification. The method enables interpretable and scalable analysis of brain MRI data, detecting disease-relevant regions through a structured low-rank factorization. By incorporating a second spatially shifted view of the data, the cyclic-shift logistic SKPD improves robustness to misalignment across subjects, a common challenge in neuroimaging. We provide asymptotic consistency guarantees under a restricted isometry condition adapted to logistic loss. Simulations confirm the model's ability to recover spatial signals under noise and identify optimal patch sizes for factor decomposition. Application to OASIS-1 and ADNI-1 datasets demonstrates that the model achieves strong classification accuracy and localizes estimated coefficients in clinically relevant brain regions, such as the hippocampus. A data-driven slice selection strategy further improves interpretability in 2D projections. The proposed framework offers a principled, interpretable, and computationally efficient tool for neuroimaging-based disease diagnosis, with potential extensions to multi-class settings and more complex transformations.

stat.ME↗