English

An interpretable neural network-based non-proportional odds model for ordinal regression

Methodology 2024-03-13 v4 Machine Learning Machine Learning

Abstract

This study proposes an interpretable neural network-based non-proportional odds model (N3^3POM) for ordinal regression. N3^3POM is different from conventional approaches to ordinal regression with non-proportional models in several ways: (1) N3^3POM is defined for both continuous and discrete responses, whereas standard methods typically treat the ordered continuous variables as if they are discrete, (2) instead of estimating response-dependent finite-dimensional coefficients of linear models from discrete responses as is done in conventional approaches, we train a non-linear neural network to serve as a coefficient function. Thanks to the neural network, N3^3POM offers flexibility while preserving the interpretability of conventional ordinal regression. We establish a sufficient condition under which the predicted conditional cumulative probability locally satisfies the monotonicity constraint over a user-specified region in the covariate space. Additionally, we provide a monotonicity-preserving stochastic (MPS) algorithm for effectively training the neural network. We apply N3^3POM to several real-world datasets.

Keywords

Cite

@article{arxiv.2303.17823,
  title  = {An interpretable neural network-based non-proportional odds model for ordinal regression},
  author = {Akifumi Okuno and Kazuharu Harada},
  journal= {arXiv preprint arXiv:2303.17823},
  year   = {2024}
}

Comments

35 pages, 18 figures, accepted to Journal of Computational and Graphical Statistics

R2 v1 2026-06-28T09:42:30.577Z