Nicolas Lanzetti

Last Name

Lanzetti

First Name

Nicolas

ORCID

0000-0002-9128-1412

Organisational unit

09478 - Dörfler, Florian / Dörfler, Florian

Show all metadata

Search Results

Publications 1 - 10 of 26

Dynamic Programming in Probability Spaces via Optimal Transport
Terpin, Antonio; Lanzetti, Nicolas; Dörfler, Florian (2024)
SIAM Journal on Control and Optimization
We study discrete-time finite-horizon optimal control problems in probability spaces, whereby the state of the system is a probability measure. We show that, in many instances, the solution of dynamic programming in probability spaces results from two ingredients: (i) the solution of dynamic programming in the “ground space” (i.e., the space on which the probability measures live) and (ii) the solution of an optimal transport problem. From a multi-agent control perspective, a separation principle holds: “low-level control of the agents of the fleet” (how does one reach the destination?) and “fleet-level control” (who goes where?) are decoupled.
Gradient-flow SDEs have unique transient population dynamics
Guan, Vincent; Janssen, Joseph; Lanzetti, Nicolas; et al. (2025)
arXiv
Identifying the drift and diffusion of an SDE from its population dynamics is a notoriously challenging task. Researchers in machine learning and single cell biology have only been able to prove a partial identifiability result: for potential-driven SDEs, the gradient-flow drift can be identified from temporal marginals if the Brownian diffusivity is already known. Existing methods therefore assume that the diffusivity is known a priori, despite it being unknown in practice. We dispel the need for this assumption by providing a complete characterization of identifiability: the gradient-flow drift and Brownian diffusivity are jointly identifiable from temporal marginals if and only if the process is observed outside of equilibrium. Given this fundamental result, we propose nn-APPEX, the first Schrödinger Bridge-based inference method that can simultaneously learn the drift and diffusion of gradient-flow SDEs solely from observed marginals. Extensive numerical experiments show that nn-APPEX's ability to adjust its diffusion estimate enables accurate inference, while previous Schrödinger Bridge methods obtain biased drift estimates due to their assumed, and likely incorrect, diffusion.
Trust Region Policy Optimization with Optimal Transport Discrepancies: Duality and Algorithm for Continuous Actions
Terpin, Antonio; Lanzetti, Nicolas; Yardim, Batuhan; et al. (2022)
arXiv
Policy Optimization (PO) algorithms have been proven particularly suited to handle the high-dimensionality of real-world continuous control tasks. In this context, Trust Region Policy Optimization methods represent a popular approach to stabilize the policy updates. These usually rely on the Kullback-Leibler (KL) divergence to limit the change in the policy. The Wasserstein distance represents a natural alternative, in place of the KL divergence, to define trust regions or to regularize the objective function. However, state-of-the-art works either resort to its approximations or do not provide an algorithm for continuous state-action spaces, reducing the applicability of the method. In this paper, we explore optimal transport discrepancies (which include the Wasserstein distance) to define trust regions, and we propose a novel algorithm - Optimal Transport Trust Region Policy Optimization (OT-TRPO) - for continuous state-action spaces. We circumvent the infinite-dimensional optimization problem for PO by providing a one-dimensional dual reformulation for which strong duality holds. We then analytically derive the optimal policy update given the solution of the dual problem. This way, we bypass the computation of optimal transport costs and of optimal transport maps, which we implicitly characterize by solving the dual formulation. Finally, we provide an experimental evaluation of our approach across various control tasks. Our results show that optimal transport discrepancies can offer an advantage over state-of-the-art approaches.
Trust Region Policy Optimization with Optimal Transport Discrepancies: Duality and Algorithm for Continuous Actions
Terpin, Antonio; Lanzetti, Nicolas; Yardim, Batuhan; et al. (2022)
Advances in Neural Information Processing Systems 35
Policy Optimization (PO) algorithms have been proven particularly suited to handle the high-dimensionality of real-world continuous control tasks. In this context, Trust Region Policy Optimization methods represent a popular approach to stabilize the policy updates. These usually rely on the Kullback-Leibler (KL) divergence to limit the change in the policy. The Wasserstein distance represents a natural alternative, in place of the KL divergence, to define trust regions or to regularize the objective function. However, state-of-the-art works either resort to its approximations or do not provide an algorithm for continuous state-action spaces, reducing the applicability of the method.In this paper, we explore optimal transport discrepancies (which include the Wasserstein distance) to define trust regions, and we propose a novel algorithm - Optimal Transport Trust Region Policy Optimization (OT-TRPO) - for continuous state-action spaces. We circumvent the infinite-dimensional optimization problem for PO by providing a one-dimensional dual reformulation for which strong duality holds.We then analytically derive the optimal policy update given the solution of the dual problem. This way, we bypass the computation of optimal transport costs and of optimal transport maps, which we implicitly characterize by solving the dual formulation.Finally, we provide an experimental evaluation of our approach across various control tasks. Our results show that optimal transport discrepancies can offer an advantage over state-of-the-art approaches.
Optimality of Linear Policies for Distributionally Robust Linear Quadratic Gaussian Regulator with Stationary Distributions
Lanzetti, Nicolas; Terpin, Antonio; Dörfler, Florian (2024)
arXiv
We prove that linear policies remain optimal for solving the Linear Quadratic Gaussian regulation problem in face of worst-case process and measurement noise distributions, when these are independent, stationary, and known to be within a radius (in the Wasserstein sense) to some reference Gaussian noise distributions. When the reference noise distributions are not Gaussian, we provide a closed-form solution for the worst-case distributions. Our main result suggests a computational complexity that scales only with the dimensionality of the system, and provides a less conservative alternative to recent work in distributionally robust control.
Variational Analysis in the Wasserstein Space
Lanzetti, Nicolas; Terpin, Antonio; Dörfler, Florian (2024)
arXiv
We study optimization problems whereby the optimization variable is a probability measure. Since the probability space is not a vector space, many classical and powerful methods for optimization (e.g., gradients) are of little help. Thus, one typically resorts to the abstract machinery of infinite-dimensional analysis or other ad-hoc methodologies, not tailored to the probability space, which however involve projections or rely on convexity-type assumptions. We believe instead that these problems call for a comprehensive methodological framework for calculus in probability spaces. In this work, we combine ideas from optimal transport, variational analysis, and Wasserstein gradient flows to equip the Wasserstein space (i.e., the space of probability measures endowed with the Wasserstein distance) with a variational structure, both by combining and extending existing results and introducing novel tools. Our theoretical analysis culminates in very general necessary optimality conditions for optimality. Notably, our conditions (i) resemble the rationales of Euclidean spaces, such as the Karush-Kuhn-Tucker and Lagrange conditions, (ii) are intuitive, informative, and easy to study, and (iii) yield closed-form solutions or can be used to design computationally attractive algorithms. We believe this framework lays the foundation for new algorithmic and theoretical advancements in the study of optimization problems in probability spaces, which we exemplify with numerous case studies and applications to machine learning, drug discovery, and distributionally robust optimization.
Learning diffusion at lightspeed
Terpin, Antonio; Lanzetti, Nicolas; Gadea, Martín; et al. (2024)
Advances in Neural Information Processing Systems 37
Diffusion regulates numerous natural processes and the dynamics of many successful generative models. Existing models to learn the diffusion terms from observational data rely on complex bilevel optimization problems and model only the drift of the system. We propose a new simple model, JKOnet*, which bypasses the complexity of existing architectures while presenting significantly enhanced representational capabilities: JKOnet* recovers the potential, interaction, and internal energy components of the underlying diffusion process. JKOnet* minimizes a simple quadratic loss and outperforms other baselines in terms of sample efficiency, computational complexity, and accuracy. Additionally, JKOnet* provides a closed-form optimal solution for linearly parametrized functionals, and, when applied to predict the evolution of cellular processes from real-world data, it achieves state-of-the-art accuracy at a fraction of the computational cost of all existing methods. Our methodology is based on the interpretation of diffusion processes as energy-minimizing trajectories in the probability space via the so-called JKO scheme, which we study via its first-order optimality conditions.
Modeling of Political Systems using Wasserstein Gradient Flows
Lanzetti, Nicolas; Hajar, Joudi; Dörfler, Florian (2022)
arXiv
The study of complex political phenomena such as parties' polarization calls for mathematical models of political systems. In this paper, we aim at modeling the time evolution of a political system whereby various parties selfishly interact to maximize their political success (e.g., number of votes). More specifically, we identify the ideology of a party as a probability distribution over a one-dimensional real-valued ideology space, and we formulate a gradient flow in the probability space (also called a Wasserstein gradient flow) to study its temporal evolution. We characterize the equilibria of the arising dynamic system, and establish local convergence under mild assumptions. We calibrate and validate our model with real-world time-series data of the time evolution of the ideologies of the Republican and Democratic parties in the US Congress. Our framework allows to rigorously reason about various political effects such as parties' polarization and homogeneity. Among others, our mechanistic model can explain why political parties become more polarized and less inclusive with time (their distributions get "tighter"), until all candidates in a party converge asymptotically to the same ideological position.
First-Order Conditions for Optimization in the Wasserstein Space
Lanzetti, Nicolas; Bolognani, Saverio; Dörfler, Florian (2025)
SIAM Journal on Mathematics of Data Science
We study first-order optimality conditions for constrained optimization in the Wasserstein space, whereby one seeks to minimize a real-valued function over the space of probability measures endowed with the Wasserstein distance. Our analysis combines recent insights on the geometry and the differential structure of the Wasserstein space with more classical calculus of variations. We show that simple rationales such as “setting the derivative to zero” and “gradients are aligned at optimality” carry over to the Wasserstein space. We deploy our tools to study and solve optimization problems in the setting of distributionally robust optimization and statistical inference. The generality of our methodology allows us to naturally deal with functionals, such as mean-variance, Kullback–Leibler divergence, and Wasserstein distance, which are traditionally difficult to study in a unified framework.
Modeling of Political Systems using Wasserstein Gradient Flows
Lanzetti, Nicolas; Hajar, Joudi; Dörfler, Florian (2022)
2022 IEEE 61st Conference on Decision and Control (CDC)
The study of complex political phenomena such as parties’ polarization calls for mathematical models of political systems. In this paper, we aim at modeling the time evolution of a political system whereby various parties selfishly interact to maximize their political success (e.g., number of votes). More specifically, we identify the ideology of a party as a probability distribution over a one-dimensional real-valued ideology space, and we formulate a gradient flow in the probability space (also called a Wasserstein gradient flow) to study its temporal evolution. We characterize the equilibria of the arising dynamic system, and establish local convergence under mild assumptions. We calibrate and validate our model with real-world time-series data of the time evolution of the ideologies of the Republican and Democratic parties in the US Congress. Our framework allows to rigorously reason about various political effects such as parties’ polarization and homogeneity. Among others, our mechanistic model can explain why political parties become more polarized and less inclusive with time (their distributions get "tighter"), until all candidates in a party converge asymptotically to the same ideological position.

Publications 1 - 10 of 26

Nicolas Lanzetti

Last Name

First Name

ORCID

Organisational unit

Filters

Search Results