Mengdi Wang

Mengdi Wang is a theoretical computer scientist who is a professor at Princeton University. Her research considers the fundamental theory that underpins reinforcement and machine learning. She was named one of MIT Technology Review 's 35 Under 35 in 2018.

Early life and education
Wang was an undergraduate student at Tsinghua University, where she specialized in automation. At the age of 18, she joined Massachusetts Institute of Technology as a graduate student, where she worked alongside Dimitri Bertsekas. Her doctoral research developed stochastic methods for large-scale linear systems.

Research and career
Wang specializes in the theoretical frameworks that underpin machine learning and reinforcement learning. She joined Princeton University as an assistant professor in 2014. She was the first person to propose stochastic gradient methods for composition optimisation. Her early work used reinforcement to minimize risk in financial portfolios and help hospitals identify potential complications.

Wang has studied Markov decision processes, a model for reinforcement learning. She uses state compression methods to use empirical data to sketch black box Markov processes.

In 2020, Wang joined the C3.ai Digital Transformation Institute, a consortium of researchers who seek to accelerate the use of artificial intelligence in society. She proposed that reinforcement learning could be used to protect educational establishments from COVID-19. She used system identification and adaptive control to develop strategies to understand the health status of students, and to deploy algorithms that recommend interventions to decision makers. In 2024, she was awarded a United States Department of Defense Multidisciplinary University Research Initiative program to develop AI and reinforcement learning for biological systems. She showed it was possible to use large language models with semantic representation to design MRNA vaccines.

Awards and honors

 * 2016 Mathematical Optimization Society Young Researcher Prize in Continuous Optimization
 * 2016 Princeton SEAS Innovation Award
 * 2017 NSF Career Award
 * 2017 Google Faculty Award
 * 2018 MIT Tech Review 35-Under-35
 * 2022 WAIC YunFan Award
 * 2024 American Automatic Control Council Donald Eckman Award