Beyond Box-Cox: A diffusion-inspired functional framework for nonlinear demand and discrete choice modeling
Year: 2025
Type: article
Abstract: This paper introduces a diffusion-inspired functional framework for modeling nonlinear demand and discrete choice in transportation and related fields. The approach extends beyond traditional Box-Cox transformations by integrating a diffusion-type function with a bounded transfer function. The resulting formulation ensures monotonicity, concavity, and convexity through explicit, verifiable conditions on its components. This systematic framework offers a tractable, low-parameter, and modular way to model demand, pricing responses, and consumer choices. The flexibility and advantages are demonstrated in two case studies: discrete choice for mode selection and logistic regression for vehicle ownership. These applications show how the method improves model fit, offers better control over tail behavior, and refines elasticity estimates. Controlled simulations further confirm its robustness across diverse nonlinear processes. By allowing researchers to integrate various functional forms, our methodology enhances interpretability, computational efficiency, and ease of estimation in transportation and related fields.
Author: Jeppe Rich
Institution: Technical University of Denmark
Language: English
Subfield: Transportation
Field: Social Sciences
Domain: Social Sciences
Open Access status: hybrid
Estimated APC paid: $3,810
OpenAlex ID: W7116095035