Selected work

Representative Papers

A Bayesian Discrete Framework for Enhancing Decision-Making Processes in Clinical Trial Designs and Evaluations

Jan 15, 2026

This study addresses the limitations of conventional frequentist approaches in effectively incorporating prior knowledge, which constrains adaptive decision-making and reliability in clinical trials. The authors propose a Bayesian framework tailored for discrete probability distributions—such as binomial, Poisson, and negative binomial—to model binary responses and overdispersed clinical endpoints using Bayesian networks. By continuously integrating accumulating evidence, the framework dynamically optimizes trial design and evaluation. Compared to maximum likelihood estimation, this approach demonstrates greater flexibility and robustness in both inferential behavior and practical performance, substantially enhancing decision quality while mitigating misinterpretation of results and reproducibility challenges.

1 citationsRead paper

Bayesian Signaling and Entry Decisions under Uncertain Market Conditions

Aug 17, 2026

本文通过结合CKLS需求不确定性、私人信息、不可逆进入决策等要素，利用贝叶斯信号传递和路径积分控制方法解决市场条件不确定下的进入决策问题。

0 citationsRead paper

Dynamic Physical Hedging amid Jump Losses, Reconstruction-Price Uncertainty, Population Interactions

Aug 13, 2026

This study addresses dynamic physical hedging under the joint risks of catastrophe losses and stochastic reconstruction costs by formulating a controlled jump-diffusion mean-field game model. We establish the viscosity solution and relaxed equilibrium theory for the associated non-local HJB-Kolmogorov coupled system, proving the existence and uniqueness of the equilibrium. The analysis elucidates the intrinsic mechanisms through which reconstruction cost uncertainty and population vulnerability influence optimal strategies. Furthermore, numerical simulations demonstrate that capitalization levels significantly alter hedging behaviors and confirm the tail robustness of the proposed strategies. Collectively, this work provides a novel theoretical framework for insurance hedging within complex risk environments characterized by systemic interactions and discontinuous dynamics.

0 citationsRead paper

Stochastic Choice with Distribution-Dependent Preferences

Aug 06, 2026

This study addresses the inability of traditional dynamic random utility models to capture endogenous feedback from behavior to preferences. It proposes a continuous-time stochastic choice theory that formalizes this feedback through the joint distribution of latent preferences and observable choices: current actions influence the evolution of future preferences via conditional distributions. The work innovatively introduces a behavioral representation in which preferences depend on the distribution of past choices, thereby establishing a rigid link between preference dynamics and observed behavior. It further demonstrates that, in the presence of such distributional feedback, the standard dynamic random utility representation no longer holds. By leveraging a conditional McKean–Vlasov system and integrating behavioral identification with continuous-time stochastic process analysis, the paper establishes existence and weak uniqueness of the model solution, providing a unified framework for identifying endogenous information, preference evolution, and stochastic choice.

0 citationsRead paper

Modeling Educational Performance Using School Demographics and Teacher Characteristics

Jun 25, 2026

High-dimensional educational data often exhibit sparsity, grouped predictor structures, and local correlations, which limit the performance of traditional regression methods. This work proposes an adaptive weighted group-fused LASSO estimator that simultaneously achieves adaptive variable selection, group-wise sparsity regularization, and coefficient fusion within a unified penalized regression framework. An efficient ADMM algorithm is developed for computation. The method uniquely integrates these three components and enjoys strong theoretical guarantees, including model selection consistency, the oracle property, and debiased asymptotic normality. Empirical results demonstrate its superior estimation accuracy and predictive performance over existing penalized approaches. Applied to Alabama state mathematics achievement data, the proposed method significantly enhances model interpretability and prediction accuracy while effectively identifying key school-level determinants.

0 citationsRead paper

Recent publications

Latest Papers

Bayesian Signaling and Entry Decisions under Uncertain Market Conditions

Aug 17, 2026

本文通过结合CKLS需求不确定性、私人信息、不可逆进入决策等要素，利用贝叶斯信号传递和路径积分控制方法解决市场条件不确定下的进入决策问题。

0 citationsRead paper

Dynamic Physical Hedging amid Jump Losses, Reconstruction-Price Uncertainty, Population Interactions

Aug 13, 2026

This study addresses dynamic physical hedging under the joint risks of catastrophe losses and stochastic reconstruction costs by formulating a controlled jump-diffusion mean-field game model. We establish the viscosity solution and relaxed equilibrium theory for the associated non-local HJB-Kolmogorov coupled system, proving the existence and uniqueness of the equilibrium. The analysis elucidates the intrinsic mechanisms through which reconstruction cost uncertainty and population vulnerability influence optimal strategies. Furthermore, numerical simulations demonstrate that capitalization levels significantly alter hedging behaviors and confirm the tail robustness of the proposed strategies. Collectively, this work provides a novel theoretical framework for insurance hedging within complex risk environments characterized by systemic interactions and discontinuous dynamics.

0 citationsRead paper

Stochastic Choice with Distribution-Dependent Preferences

Aug 06, 2026

This study addresses the inability of traditional dynamic random utility models to capture endogenous feedback from behavior to preferences. It proposes a continuous-time stochastic choice theory that formalizes this feedback through the joint distribution of latent preferences and observable choices: current actions influence the evolution of future preferences via conditional distributions. The work innovatively introduces a behavioral representation in which preferences depend on the distribution of past choices, thereby establishing a rigid link between preference dynamics and observed behavior. It further demonstrates that, in the presence of such distributional feedback, the standard dynamic random utility representation no longer holds. By leveraging a conditional McKean–Vlasov system and integrating behavioral identification with continuous-time stochastic process analysis, the paper establishes existence and weak uniqueness of the model solution, providing a unified framework for identifying endogenous information, preference evolution, and stochastic choice.

0 citationsRead paper

Modeling Educational Performance Using School Demographics and Teacher Characteristics

Jun 25, 2026

High-dimensional educational data often exhibit sparsity, grouped predictor structures, and local correlations, which limit the performance of traditional regression methods. This work proposes an adaptive weighted group-fused LASSO estimator that simultaneously achieves adaptive variable selection, group-wise sparsity regularization, and coefficient fusion within a unified penalized regression framework. An efficient ADMM algorithm is developed for computation. The method uniquely integrates these three components and enjoys strong theoretical guarantees, including model selection consistency, the oracle property, and debiased asymptotic normality. Empirical results demonstrate its superior estimation accuracy and predictive performance over existing penalized approaches. Applied to Alabama state mathematics achievement data, the proposed method significantly enhances model interpretability and prediction accuracy while effectively identifying key school-level determinants.

0 citationsRead paper

New Confidence Regions for Linear Regression Parameters with Stationary-Ergodic Dependent Errors

May 19, 2026

This study addresses the challenge of constructing valid joint confidence regions for linear regression coefficients when regression errors exhibit unknown serial dependence and are jointly stationary and ergodic with the covariates. The authors propose a novel approach that avoids explicit modeling of the error dependence structure by introducing independent auxiliary samples and applying stochastic smoothing with a decaying bandwidth to both the regression function and second moments. Coupled with data-driven bandwidth selection and mild truncation, this method yields Wald-type confidence regions and simultaneous confidence intervals. It does not rely on long-run variance estimation or parametric assumptions about dependence, achieving coverage probabilities close to nominal levels across diverse dependence structures—including ARMA, ARFIMA, copula-based Markov processes, and fractional Gaussian noise—while producing smaller confidence region volumes than Newey–West HAC and MAC methods. The approach is successfully demonstrated in an analysis of Beijing PM2.5 data.

0 citationsRead paper