no code implementations • 22 Apr 2024 • Rajiv Sambharya, Bartolomeo Stellato
We build generalization guarantees for classical optimizers, using a sample convergence bound, and for learned optimizers, using the Probably Approximately Correct (PAC)-Bayes framework.
no code implementations • 2 Apr 2024 • Filippo Fabiani, Bartolomeo Stellato, Daniele Masti, Paul J. Goulart
We consider the problem of designing a machine learning-based model of an unknown dynamical system from a finite number of (state-input)-successor state data points, such that the model obtained is also suitable for optimal control design.
no code implementations • 14 Feb 2024 • Haimin Hu, Gabriele Dragotto, Zixu Zhang, Kaiqu Liang, Bartolomeo Stellato, Jaime F. Fisac
To solve the problem, we introduce Branch and Play (B&P), an efficient and exact algorithm that provably converges to a socially optimal order of play and its Stackelberg equilibrium.
no code implementations • 6 Nov 2023 • Gabriele Dragotto, Stefan Clarke, Jaime Fernández Fisac, Bartolomeo Stellato
We consider the problem of solving a family of parametric mixed-integer linear optimization problems where some entries in the input data change.
2 code implementations • 14 Sep 2023 • Rajiv Sambharya, Georgina Hall, Brandon Amos, Bartolomeo Stellato
We introduce a machine-learning framework to warm-start fixed-point optimization algorithms.
no code implementations • NeurIPS 2021 • Tim Seyde, Igor Gilitschenski, Wilko Schwarting, Bartolomeo Stellato, Martin Riedmiller, Markus Wulfmeier, Daniela Rus
Reinforcement learning (RL) for continuous control typically employs distributions whose support covers the entire action space.
1 code implementation • NeurIPS 2021 • Jeffrey Ichnowski, Paras Jain, Bartolomeo Stellato, Goran Banjac, Michael Luo, Francesco Borrelli, Joseph E. Gonzalez, Ion Stoica, Ken Goldberg
First-order methods for quadratic optimization such as OSQP are widely used for large-scale machine learning and embedded optimal control, where many related problems must be rapidly solved.
3 code implementations • 9 Nov 2020 • Shuvomoy Das Gupta, Bartolomeo Stellato, Bart P. G. Van Parys
A wide range of nonconvex optimization problems have this structure including (but not limited to) sparse and low-rank optimization problems.
Optimization and Control
no code implementations • NeurIPS Workshop LMCA 2020 • Abhishek Cauligi, Preston Culbertson, Mac Schwager, Bartolomeo Stellato, Marco Pavone
Mixed-integer convex programming (MICP) is a popular modeling framework for solving discrete and combinatorial optimization problems arising in various settings.
no code implementations • 30 Jun 2020 • Dimitris Bertsimas, Léonard Boussioux, Ryan Cory Wright, Arthur Delarue, Vassilis Digalakis Jr., Alexandre Jacquillat, Driss Lahlou Kitane, Galit Lukin, Michael Lingzhi Li, Luca Mingardi, Omid Nohadani, Agni Orfanoudaki, Theodore Papalexopoulos, Ivan Paskov, Jean Pauphilet, Omar Skali Lami, Bartolomeo Stellato, Hamza Tazi Bouardi, Kimberly Villalobos Carballo, Holly Wiberg, Cynthia Zeng
Specifically, we propose a comprehensive data-driven approach to understand the clinical characteristics of COVID-19, predict its mortality, forecast its evolution, and ultimately alleviate its impact.
no code implementations • L4DC 2020 • Akshay Agrawal, Shane Barratt, Stephen Boyd, Bartolomeo Stellato
Common examples of such convex optimization control policies (COCPs) include the linear quadratic regulator (LQR), convex model predictive control (MPC), and convex control-Lyapunov or approximate dynamic programming (ADP) policies.
1 code implementation • 4 Jul 2019 • Dimitris Bertsimas, Bartolomeo Stellato
Compared to state-of-the-art MIO routines, the online running time of our method is very predictable and can be lower than a single matrix factorization time.
1 code implementation • 24 Dec 2018 • Dimitris Bertsimas, Bartolomeo Stellato
We introduce the idea that using optimal classification trees (OCTs) and optimal classification trees with-hyperplanes (OCT-Hs), interpretable machine learning algorithms developed by Bertsimas and Dunn [2017, 2018], we are able to obtain insight on the strategy behind the optimal solution in continuous and mixed-integer convex optimization problem as a function of key parameters that affect the problem.
Optimization and Control
2 code implementations • 21 Nov 2017 • Bartolomeo Stellato, Goran Banjac, Paul Goulart, Alberto Bemporad, Stephen Boyd
We present a general purpose solver for convex quadratic programs based on the alternating direction method of multipliers, employing a novel operator splitting technique that requires the solution of a quasi-definite linear system with the same coefficient matrix at almost every iteration.
Optimization and Control
no code implementations • 19 Oct 2015 • Bartolomeo Stellato, Tobias Geyer, Paul J. Goulart
To the authors' knowledge, this is the first time direct MPC for current control has been implemented on an FPGA solving the integer optimization problem in real-time and achieving comparable performance to formulations with long prediction horizons.
Optimization and Control