Peng Liao 廖鹏

About Me

I am currently a Quantitative Researcher at DRW. Prior to joining DRW in October 2021, I was a postdoctoral research fellow in the Department of Statistics at Harvard University under supervision of Prof. Susan Murphy. I received Doctoral degree in Statistics from the University of Michigan and Bachelor’s degree in Statistics from Sun Yat-Sen University (Zhongshan University).

My profile at Google Scholar.

Research Interest

My research focuses on sequential decision-making problems with applications in mobile health. In particular, I work on developing

  • new experimental designs (micro-randomized trial, MRT) for use in mobile health

  • off-line/batch data analysis methods using data collected from MRT (causal inference, policy evaluation/learning)

  • online Reinforcement Learning algorithms that sequentially select interventions based on past history with the goal to maximize the long-term outcomes.

Education

  • 2014 - 2019: Ph.D in Statistics, University of Michigan, Ann Arbor, MI, United States.

  • 2009 - 2013: B.S. in Statistics, School of Mathematics and Computational Science, Sun Yat-Sen University, Guangzhou, P.R.China.

Selected Publications

[1] Liao, P., Klasnja, P., Tewari, A. and Murphy, S.A., 2016. Sample size calculations for micro-randomized trials in mHealth. Statistics in medicine, 35(12), pp.1944-1971. (An online sample size calculator could be assessed here (see more). The online calculator may not be accessible due to high traffic. Please download the R package instead.)

[2] Liao, P.*, Dempsey, W.*, Sarker, H., Hossain, S.M., al'Absi, M., Klasnja, P. and Murphy, S., 2018. Just-in-Time but Not Too Much: Determining Treatment Timing Mobile Health. Proceedings of the ACM on Interactive, Mobile, Wearable and Ubiquitous Technologies (IMWUT), 2(4), p.179. (*equal contribution)

[3] Dempsey, W., Liao, P., Kumar, S. and Murphy, S.A., 2020. The stratified micro-randomized trial design: sample size considerations for testing nested causal effects of time-varying treatments. Annals of Applied Statistics, Vol. 14, No.2, 661-684.

[4] Liao, P., Greenewald, K., Klasnja,P., and Murphy, S.A., 2020. PersonalizedHeartSteps: A Reinforcement Learning Algorithm for Optimizing PhysicalActivity. Proceedings of the ACM on Interactive, Mobile, Wearable and Ubiquitous Technologies (IMWUT), 2020 Mar;4(1):18.

[5] Liao, P., Klasnja, P., and Murphy, S.A., 2021. Off-Policy Estimation of Long-Term Average Outcomes with Applications to Mobile Health. Journal of the American Statistical Association, 116:533, 382-391

[6] Tomkins, S., Liao, P., Klasnja, P. and Murphy, S., 2021. IntelligentPooling: Practical Thompson Sampling for mHealth. Journal of Machine Learning, 110:2685–2727.

[7] Liao, P., Qi, Z. and Murphy, S.A., (2022+). Batch Policy Learning in Average Reward Markov Decision Processes. To appear in Annals of Statistics.

Work in Preparation

[1] Qi, Z. and Liao, P. Robust Batch Policy Learning in Markov Decision Processes. Submitted to Operation Research (under review). arXiv link.

Contact

  • E-mail: pliao9122 at gmail dot com