Institution profile

Oklahoma State University

Academic institutionnorthamerica · us
Official website
Research library73linked papers
Opportunities0open roles
Selected work

Representative Papers

Evaluating the Impact of COVID-19 on Transportation Infrastructure Funding in the United States

Aug 31, 2022International Conference on Transportation and Development 2022

This study addresses the significant decline in motor fuel consumption during the COVID-19 pandemic and its adverse impact on U.S. state-level transportation infrastructure funding, which heavily relies on fuel tax revenues. For the first time, the research integrates pandemic dynamics, fuel consumption patterns, and demographic data into a machine learning framework to quantify the long-term fiscal effects on state transportation budgets and forecast future trends. The proposed model demonstrates exceptional predictive performance at the state level (R² > 95%), accurately capturing fluctuations in fuel usage. Projections indicate that fuel tax revenues in several states are expected to remain 10%–15% below pre-pandemic levels for one to two years, providing policymakers with a data-driven foundation for strategic planning and fiscal adaptation.

2 citationsRead paper

Vertical tacit collusion in AI-mediated markets

Jan 06, 2026arXiv.org

The proliferation of AI-powered shopping agents has given rise to a novel form of market failure: although platforms and sellers exhibit no explicit collusion, they jointly exploit cognitive biases inherent in large language models to the detriment of consumers. This work introduces the concept of “vertical tacit collusion” and develops a multi-agent reinforcement learning simulation environment calibrated with empirical data on LLM biases to model the strategic interaction between platform-controlled product rankings and seller-manipulated product descriptions. The findings reveal that the combined effect of these strategies inflicts more than twice the consumer harm compared to the sum of their independent effects, uncovering an AI-driven mechanism of anticompetitive harm that evades detection under current antitrust frameworks and exposing a critical blind spot in regulatory oversight.

1 citationsRead paper

Design-Assisted Regression

Sep 11, 2026

本文提出一种设计辅助回归框架,通过利用协变量分布信息来稳定弱设计方向和修正潜在效应扭曲,从而改进估计性能。

0 citationsRead paper
Recent publications

Latest Papers

Design-Assisted Regression

Sep 11, 2026

本文提出一种设计辅助回归框架,通过利用协变量分布信息来稳定弱设计方向和修正潜在效应扭曲,从而改进估计性能。

0 citationsRead paper