arXiv research

The paper proves optimal smoothness for certain Lagrangian graphs with specific Hölder continuity.

Researchers develop neural optimal transport with Lagrangian costs for efficient computation.

New Lagrangian approach for optimal control of second-order systems.

Study shows how to preserve Lagrangian condition in mean curvature flow on Kim-McCann metrics.

In recent years, constrained optimization has become increasingly relevant to the machine learning community, with applications including Neyman-Pearson classification, robust optimization, and fair machine learning. A natural approach to constrained optimization is to optimize the Lagrangian, but this is not guarantee…

New curvature measure for optimal transport with specific cost function.

Unified Lagrangian-based methods for nonsmooth nonconvex optimization.

This paper explores the potential of Lagrangian duality for learning applications that feature complex constraints. Such constraints arise in many science and engineering domains, where the task amounts to learning optimization problems which must be solved repeatedly and include hard physical and operational constrain…

Classifies regularity for Lagrangian mean curvature type equations.

New method solves constrained optimization problems efficiently.

A new method tackles nonconvex optimization with penalties and proximal terms.

A new method solves complex constrained minimax problems.

We prove an optimal control on the time-dependent measure of a measurable set under a reparametrized Lagrangian mean curvature flow of almost calibrated submanifolds in a Calabi-Yau manifold. Moreover we give a classification of those Lagrangian translating solitons in $\mathbb{C}^m$ that evolve by this reparametrized …

Study uses DRL with Lagrangian relaxation to solve temporal control tasks with STL constraints.

Proves smoothness and estimates for special Lagrangian solutions with semi-convexity.

A new method solves distributed optimization problems over networks.

Improves neural network verification by merging abstract domains and Lagrangian methods.

The study finds conditions for free boundary Hamiltonian stationary discs in complex 2-space.

Develops a new solver for optimizing with stochastic dominance constraints.

A large number of objectives have been proposed to train latent variable generative models. We show that many of them are Lagrangian dual functions of the same primal optimization problem. The primal problem optimizes the mutual information between latent and visible variables, subject to the constraints of accurately …

Unified deep learning framework solves various optimal transport problems.

We geometrically describe optimal control problems in terms of Morse families in the Hamiltonian framework. These geometric structures allow us to recover the classical first order necessary conditions for optimality and the starting point to run an integrability algorithm. Moreover the integrability algorithm is adapt…

Let $M$ be an $n$-dimensional Lagrangian submanifold of a complex space form. We prove a pointwise inequality $δ(n_1,\ldots,n_k) \leq a(n,k,n_1,\ldots,n_k) \|H\|^2 + b(n,k,n_1,\ldots,n_k)c,$ with on the left hand side any delta-invariant of the Riemannian manifold $M$ and on the right hand side a linear combination o…

New methods for efficient exploration under unknown linear constraints in bandits.

PDCA algorithm learns policies for RL with constraints using a primal-dual approach.

Paper proposes distributed optimization for federated learning with theoretical guarantees.

The information bottleneck (IB) problem tackles the issue of obtaining relevant compressed representations $T$ of some random variable $X$ for the task of predicting $Y$. It is defined as a constrained optimization problem which maximizes the information the representation has about the task, $I(T;Y)$, while ensuring t…

It was proved in [8,9] that every Lagrangian submanifold $M$ of a complex space form $\tilde M^{5}(4c)$ of constant holomorphic sectional curvature $4c$ satisfies the following optimal inequality: {align}\tag{A}δ(2,2)\leq \text{\small${25}{4}$} H^{2}+8c,{align} where $H^{2}$ is the squared mean curvature and $δ(2,2)$ i…

In this paper we will discuss some new developments in the design of numerical methods for optimal control problems of Lagrangian systems on Lie groups. We will construct these geometric integrators using discrete variational calculus on Lie groups, deriving a discrete version of the second-order Euler-Lagrange equatio…

Efficiently solves exploration-exploitation in LQR using Lagrangian relaxation.

The Rusk-Skinner formalism was developed in order to give a geometrical unified formalism for describing mechanical systems. It incorporates all the characteristics of Lagrangian and Hamiltonian descriptions of these systems (including dynamical equations and solutions, constraints, Legendre map, evolution operators, e…

Given a quasisymmetric homeomorphism $\varphi$ of the circle, Bonsante and Schlenker proved the existence and uniqueness of the minimal Lagrangian extension $f_\varphi:\mathbb{H}^2\to\mathbb{H}^2$ to the hyperbolic plane. By previous work of the author, its maximal dilatation satisfies $\log K(f_\varphi)\leq C||\varphi…

A new decentralized algorithm DESTINY solves optimization over Stiefel manifold with single communication round.

We study constrained nonconvex optimization problems in machine learning, signal processing, and stochastic control. It is well-known that these problems can be rewritten to a minimax problem in a Lagrangian form. However, due to the lack of convexity, their landscape is not well understood and how to find the stable e…

Accelerated gradient methods play a central role in optimization, achieving optimal rates in many settings. While many generalizations and extensions of Nesterov's original acceleration method have been proposed, it is not yet clear what is the natural scope of the acceleration concept. In this paper, we study accelera…

Bayesian optimization tackles mixed discrete-continuous problems with Gaussian processes.

We prove some non-existence theorems for translating solutions to Lagrangian mean curvature flow. More precisely, we show that translating solutions with an $L^2$ bound on the mean curvature are planes and that almost-calibrated translating solutions which are static are also planes. Recent work of D. Joyce, Y.-I. Lee,…

It was proven in [B.-Y. Chen, F. Dillen, J. Van der Veken and L. Vrancken, Curvature inequalities for Lagrangian submanifolds: the final solution, Differ. Geom. Appl. 31 (2013), 808-819] that every Lagrangian submanifold $M$ of a complex space form $\tilde M^{n}(4c)$ of constant holomorphic sectional curvature $4c$ sat…

The paper introduces a new concept of frame vorticity and uses it to find optimal sections in specific geometric settings.

The paper develops methods for time-varying constrained online convex optimization.

Unified DICE estimators as regularized Lagrangians for improved off-policy evaluation.

Paper proposes a method to find approximate SOSP for nonconvex conic optimization problems.

We address the problem of solving convex optimization problems with many convex constraints in a distributed setting. Our approach is based on an extension of the alternating direction method of multipliers (ADMM) that recently gained a lot of attention in the Big Data context. Although it has been invented decades ago…

New offline RL algorithms tackle partial data coverage with optimal performance and practicality.

Constructing translating solitons from Lagrangian Grim Reapers.

Stochastic approach improves neural network training for kinetic simulations.

We consider a multi-objective risk-averse two-stage stochastic programming problem with a multivariate convex risk measure. We suggest a convex vector optimization formulation with set-valued constraints and propose an extended version of Benson's algorithm to solve this problem. Using Lagrangian duality, we develop sc…

In this paper, we study the Lagrangian F-stability of closed Lagrangian self-shrinkers immersed in complex Euclidean space. We show that any closed Lagrangian self-shrinker with first Betti number greater than one is Lagrangian F-unstable. In particular, any two-dimensional embedded closed Lagrangian self-shrinker is L…