EUVIMEDScopri EUVIMED
Crossref

A Matrix‐Based Hidden Markov Model for Consumer Credit Analysis

Borui Qi, Lin Sun, Xiaoxia Sun

Journal of Forecasting · 2026-08 · 45 (5) · 2686-2699

DOI: 10.1002/for.70163 ↗

At a glance

Journal article in Journal of Forecasting, published online on 2026-04-15. Written by Borui Qi and 2 co-authors. 16 references.

AbstractDati bibliograficiCiteRelated studies

Abstract (original language)

This paper introduces a novel hidden Markov variant, the matrix‐based hidden Markov model (MHMM), which characterizes mixture of Markov chains whose transition matrices are governed stochastically by unobservable states. By offering a multi‐scale, hierarchical framework for stochastic dynamics, the model enables improved inference of structured hidden states (i.e., the hidden transition matrices) and their underlying generative mechanism from non‐homogeneous time series data. The paper presents efficient algorithms for parameter estimation and hidden path inference. Besides, we introduce the absorbing matrix‐based hidden Markov model (AMHMM), incorporating absorbing states into the framework. Through Markov absorption theory, we analyze the model's pre‐absorption behavior. A numerical example, based on simulated consumer repayment data, illustrates the value of the AMHMM: It provides estimates of default timing, reveals propensity for overdue installments and behavioral patterns, and helps assess the overall trend in the credit environment.

Cite

APA

Qi, B., Sun, L., & Sun, X. (2026). A Matrix‐Based Hidden Markov Model for Consumer Credit Analysis. Journal of Forecasting, 45(5), 2686-2699. https://doi.org/10.1002/for.70163

Vancouver

Qi B, Sun L, Sun X. A Matrix‐Based Hidden Markov Model for Consumer Credit Analysis. Journal of Forecasting. 2026;45(5):2686-2699. doi:10.1002/for.70163

BibTeX
@article{Qi2026,
  author = {Qi, Borui and Sun, Lin and Sun, Xiaoxia},
  title = {A Matrix‐Based Hidden Markov Model for Consumer Credit Analysis},
  journal = {Journal of Forecasting},
  year = {2026},
  volume = {45},
  number = {5},
  pages = {2686--2699},
  doi = {10.1002/for.70163}
}
Start for free →