Alexander Bodard

This work introduces an unconventional inexact augmented Lagrangian method, where the augmenting term is a Euclidean norm raised to a power between one and two. The proposed algorithm is applicable to a broad class of constrained nonconvex minimization problems, that involve nonlinear equality constraints over a convex set under a mild regularity condition. First, we conduct a full complexity analysis of the method, leveraging an accelerated first-order algorithm for solving the Ho¨lder-smooth subproblems. Next, we present an inexact proximal point method to tackle these subproblems, demonstrating that it achieves an improved convergence rate. Notably, this rate reduces to the best-known convergence rate for first-order methods when the augmenting term is a squared Euclidean norm. Our worst-case complexity results further show that using lower powers for the augmenting term leads to faster constraint satisfaction, albeit with a slower decrease in the dual residual. Numerical experiments support our theoretical findings, illustrating that this trade-off between constraint satisfaction and cost minimization is advantageous for certain practical problems.

We study generalized smoothness in nonconvex optimization, focusing on \(L_0, L_1)\-smoothness and anisotropic smoothness. The former was empirically derived from practical neural network training examples, while the latter arises naturally in the analysis of nonlinearly preconditioned gradient methods. We introduce a new sufficient condition that encompasses both notions, reveals their close connection, and holds in key applications such as phase retrieval and matrix factorization. Leveraging tools from dynamical systems theory, we then show that nonlinear preconditioning – including gradient clipping – preserves the saddle point avoidance property of classical gradient descent. Crucially, the assumptions required for this analysis are actually satisfied in these applications, unlike in classical results that rely on restrictive Lipschitz smoothness conditions. We further analyze a perturbed variant that efficiently attains second-order stationarity with only logarithmic dependence on dimension, matching similar guarantees of classical gradient methods.

In this paper we study an unconventional inexact Augmented Lagrangian Method (ALM) for convex optimization problems, as first proposed by Bertsekas, wherein the penalty term is a potentially non-Euclidean norm raised to a power between one and two. We analyze the algorithm through the lens of a nonlinear Proximal Point Method (PPM), as originally introduced by Luque, applied to the dual problem. While Luque analyzes the order of local convergence of the scheme with Euclidean norms our focus is on the non-Euclidean case which prevents us from using standard tools for the analysis such as the nonexpansiveness of the proximal mapping. To allow for errors in the primal update, we derive two implementable stopping criteria under which we analyze both the global and the local convergence rates of the algorithm. More specifically, we show that the method enjoys a fast sublinear global rate in general and a local superlinear rate under suitable growth assumptions. We also highlight that the power ALM can be interpreted as classical ALM with an implicitly defined penalty-parameter schedule, reducing its parameter dependence. Our experiments on a number of relevant problems suggest that for certain powers the method performs similarly to a classical ALM with fine-tuned adaptive penalty rule, despite involving fewer parameters.

This study introduces two second-order methods designed to provably avoid saddle points in composite nonconvex optimization problems: (i) a nonsmooth trust-region method and (ii) a curvilinear linesearch method. These developments are grounded in the forward-backward envelope (FBE), for which we analyze the local second-order differentiability around critical points and establish a novel equivalence between its second-order stationary points and those of the original objective. We show that the proposed algorithms converge to second-order stationary points of the FBE under a mild local smoothness condition on the proximal mapping of the nonsmooth term. Notably, for }( }C^2 })-partly smooth functions, this condition holds under a standard strict complementarity assumption. To the best of our knowledge, these are the first second-order algorithms that provably escape nonsmooth strict saddle points of composite nonconvex optimization, regardless of the initialization. Our preliminary numerical experiments show promising performance of the developed methods, validating our theoretical foundations.