Ludovic Tangpi
Princeton University, NJ·Mathematics & Statistics·Early-career
When many decision-makers act at once in an uncertain world, each one's best choice depends on what everyone else does. Banks lending to one another, commuters all searching for the fastest route, or people adjusting behavior during an epidemic are all examples.
- Fields
Mathematics & Statistics
Social Science
- Topics
stochastic differential games
mean field games
propagation of chaos
- Grants
- 1 known grant, through Jun 2027
What they work on
- Develop and analyze a new form of propagation of chaos for interacting particle systems that evolve backward in time, together with the associated functional inequalities, as the main tool for making the mean field game limit rigorous.
- Establish large deviation principles and non-asymptotic convergence rates to the mean field limit for both competitive and cooperative stochastic differential games, using purely probabilistic arguments.
- Apply the convergence results to numerical simulation of large-population games and to quantitative financial models of bubble formation and investment among competing agents.
- Build model-free, fully data-driven estimators of general convex risk measures using empirical process theory and propagation of chaos, and extend the techniques to large-scale optimal transportation problems.
More about the research
Mathematically these are stochastic differential games, and once the number of players gets large they become nearly impossible to solve or simulate. A long-standing shortcut says that with enough players you can replace the crowd by a single "typical" player representing the average behavior. This project aims to turn that shortcut into rigorous mathematics and to find out exactly when it works and how fast.
Why it matters
Large-population games underlie practical questions in finance, urban planning, and epidemiology, but current methods offer little guarantee that the simplifying "average player" approximation is accurate. Convergence rates and error bounds would tell practitioners how much to trust simulations of systemic risk among many banks or congestion models, and the proposed data-driven estimators could make financial risk measures computable directly from data without assuming a model.
Known grants
From NIH and NSF research grants. Labs often have other funding we can't see.
Closest labs
Asaf Cohen
University of Michigan, MI·Mathematics & Statistics
Rigorously connect the long-horizon behavior of finite-player stochastic games to their mean field game limits, identifying structural features that stay stable as the number of players grows.
Amarjit Budhiraja
University of North Carolina at Chapel Hill, NC·Mathematics & Statistics
Characterize the hydrodynamic limit of the infinite Atlas model, a rank-based particle model from stochastic portfolio theory, by linking it to Stefan-type free boundary problems, especially when particles start densely packed.
Haomin Zhou
Georgia Institute of Technology, GA·Mathematics & Statistics
Establish parameterized Wasserstein sub-manifolds, which represent evolving probability distributions through a neural network's parameters via a push-forward map and pullback metric, and develop their mathematical foundations.
The school
Princeton University: 26 labs on PI Pages, $559M a year on research, 403 research doctorates a year.
Reach out
A start for your email. Fill in the brackets.