Hengrui Cai (蔡亨瑞)

Hengrui Cai (蔡亨瑞)

Assistant Professor in Statistics

University of California Irvine

About

I am an Assistant Professor in the Department of Statistics at University of California Irvine. I obtained my Ph.D. degree in Statistics at North Carolina State University (NCSU), co-advised by Dr. Wenbin Lu and Dr. Rui Song. Prior to that, I obtained a B.S. in Statistics from Zhejiang University in July 2017.

I have broad research interests in methodology and theory in causal inference, reinforcement learning, graphical model, and their interchanges, to establish reliable, powerful, and interpretable solutions to wide real-world problems. Currently, my main research work includes individualized optimal decision making with complex data, policy evaluation in reinforcement/deep learning, and causal discovery for high-dimensional individual mediation analysis, directly motivated by precision medicine, customized economics, personalized marketing, modern epidemiology, etc.

Click here for my name in Chinese and how to pronounce it.

Download my CV . Contact me: hengrc1@uci.edu

Interests
  • Causal Inference and Causal Structure Learning
  • Reinforcement/Deep Learning
  • Policy Optimization and Evaluation
  • Precision Medicine
Education
  • PhD in Statistics, 2022

    North Carolina State University

  • B.S. in Statistics, 2017

    Zhejiang University

Awards and Honors

  • ENAR Distinguished Student Paper Award, The International Biometric Society, 2020
  • Nominated for the Outstanding TA Award, NCSU, 2019 and 2021.
  • Mu Sigma Rho, National Statistics Honor Society, NCSU, 2017 - Present.
  • National Undergraduate Research Fund, 2017
  • Meritorious Winner in American Mathematical Contest, 2016

Publications

Published/Accepted

Preprints/Under Review

(* co-first author)

Teaching

I was a teaching fellow in Department of Statistics at North Carolina State University.

Lab Instructor for Graduate Course:

  • ST 703: Statistical Methods I (Fall 2019)
  • ST 114: Introduction to Statistical Programming - Python (Spring 2022).

Teaching Fellow for Graduate Courses:

  • ST 745: Analysis of Survival Data (Spring 2021)
  • ST 405/505: Applied Nonparametric Statistics (Fall 2020)
  • ST 790: Financial Statistics (Fall 2018)
  • ST 511: Introduction to Statistics for Biological Sciences (Fall 2018).

Teaching Fellow for Undergraduate Courses:

  • ST 422: Introduction to Mathematical Statistics II (Fall 2021)
  • ST 311: Introduction to Statistics (Spring 2019)
  • ST 312: Introduction to Statistics II (Spring 2018)
  • ST 350: Economic and Business Statistics (Fall 2017).

Software

  • DJL: Deep Jump Learning for Off-Policy Evaluation in Continuous Treatment Settings. Available on GitHub.

  • ANOCE-CVAE: Analysis of Causal Effects with Multiple Mediators via Constrained Structural Learning. Available on GitHub.

  • CODA: Calibrated Optimal Decision Making with Multiple Data Sources and Limited Outcome. Available on GitHub.

  • CAPITAL: Optimal Subgroup Identification via Constrained Policy Tree Search. Available on GitHub.

  • JQL: Jump Q-Learning for Individualized Interval-Valued Dose Rule. Available on CRAN.

  • APtool: Average Positive Predictive Values (AP) for Binary Outcomes and Censored Event Times. Available on CRAN.

  • APRL: Assets Portfolio Model by Reinforcement Learning. Available on GitHub.

Recent & Upcoming Talks

  • Doubly Robust Interval Estimation for Optimal Policy Evaluation in Online Learning. The 5th International Conference on Econometrics and Statistics (EcoSta), June 2022, Kyoto, Japan. Invited.
  • Multi-Agent Graph Cooperative Bandit. NeurIPS-21 Causal Inference Challenges in Sequential Decision Making: Bridging Theory and Practice Workshop, Dec 2021, Virtual. Invited.
  • Doubly Robust Interval Estimation for Optimal Policy Evaluation in Online Learning. 2021 Computational and Methodological Statistics (CMStatistics), Dec 2021, King’s College London. Invited.
  • Deep Jump Learning for Off-Policy Evaluation in Continuous Treatment Settings. The 35th Conference on Neural Information Processing Systems (NeurIPS), Dec 2021, Virtual. Invited. Slides
  • Explainable Causal Graph Learning for Hero Feature Discovery. Explainable AI Workshop at the 9th Amazon Machine Learning Conference, Oct 2021, Virtual. Invited.
  • Calibrated Optimal Decision Making with Multiple Data Sources and Limited Outcome. Joint Statistics Meeting (JSM) 2021, Aug 2021, Virtual. Invited. Slides
  • Learning Periodic World with Gaussian Process Bandits. IJCAI-21 Reinforcement Learning for Intelligent Transportation Systems Workshop, Aug 2021, Virtual. Invited. Slides
  • Analysis of Causal Effects with Multiple Mediators via Constrained Structural Learning. The 9th International Conference on Learning Representations (ICLR), May 2021, Virtual. Invited. Slides
  • On Optimal Treatment Decision Making by Auxiliary Data with Application to AIDs Study. The Duke-Industry Statistics Symposium (DISS) 2021, April 2021, Virtual. Invited. Slides
  • Deep Jump Q-Evaluation for Offline Policy Evaluation in Continuous Action Space. The 2021 ENAR Spring Meeting, March 2021, Virtual. Student Paper Award Winner. Slides
  • A Bandit Framework for Dynamic Pricing. Reinforcement Learning for E-commerce Workshop at the 8th Amazon Machine Learning Conference, Oct 2020, Virtual. Invited.
  • Marketing Experiment Bridging: Time Inverse Bayesian Learning (TIBL). The 8th Amazon Machine Learning Conference, Sep 2020, Virtual. Invited.
  • On Validation and Planning of An Optimal Decision Rule with Application in Healthcare Studies. The 37th International Conference on Machine Learning (ICML), July 2020, Virtual. Invited. Slides

Contact