Searcharxiv⌕ Search

arXiv subjects

Felipe Lara

Publications and source records attributed to Felipe Lara.

At least 19 recordsLinked to original sources

Subgradient and Stochastic Finite-Difference Methods for Constrained Strongly Quasiconvex Optimization

We propose a projected subgradient method for solving quasiconvex minimization problems. By employing the Clarke subdifferential, we prove that the sequence generated by the proposed algorithm converges to a solution of the minimization problem when the objective function is locally Lipschitz continuous and strongly quasiconvex in the sense of Polyak, under the same assumptions required for convex functions with the convex subdifferential. Moreover, we derive explicit convergence rate estimates for both the iterates and objective function values. In particular, compared with the corresponding results for nonsmooth, locally Lipschitz, strongly convex optimization, our analysis replaces strong convexity with strong quasiconvexity while preserving the order of the convergence rate, thereby extending the guarantees to a class of possibly nonconvex objective functions. In addition, we investigate a stochastic finite-difference scheme for minimizing a strongly quasiconvex function subject to inequality constraints defined by strongly quasiconvex functions, and establish a convergence rate of order $O(k^{-1/4})$, consistent with known complexity bounds for stochastic zeroth-order methods.

math.OC↗

On Star Monotone Operators: Theory and Applications

Introduced by Solodov and Svaiter \cite{SS}, star monotone operators have repeatedly appeared as a sufficient condition ensuring global convergence of first-order algorithms, without requiring full monotonicity. Yet, despite this recurring role, a systematic study of the class itself has been missing from the literature. This paper fills that gap with the first comprehensive treatment of star monotone operators, a class that strictly encompasses monotone and pseudomonotone operators with nonempty zero sets, together with operators that are strongly quasimonotone along rays and satisfy a radial nondegeneracy condition. We develop basic calculus rules, existence and uniqueness results for zeros, global metric subregularity, and precise structural relationships with classical monotonicity notions. We further demonstrate the versatility of star monotonicity through a Clarke subdifferential characterization of nonsmooth star quasiconvexity, applications to Minty and Stampacchia variational inequalities, linear convergence guarantees for subgradient-type methods, and separable operators related to the Debreu--Koopmans problem.

math.OC↗

A Proximal Gradient Method for Nonconvex Nonsmooth Optimization with Applications

We study a proximal-gradient method for the sum of a convex smooth function $g$ and a possibly nonconvex, nonsmooth, strongly quasar-convex regularizer. The gradient step is weighted by the quasar-convexity parameter, so, for step sizes in the descent range, the iteration decreases a weighted objective, while the unweighted sum serves for comparison. We establish a geometric envelope around every global minimizer of the unweighted objective, with an explicit radius independent of the step size; a common minimizer yields global linear convergence. We also prove that every critical point $u$ of the weighted objective lies within an explicit multiple of $\|\nabla g(u)\|$ from every global minimizer of the unweighted objective. Together with an optimized composite comparison, this yields asymptotic distance bounds. An example shows that the two objectives can have different critical points. For power-regularized least squares with powers between zero and one, descent provides a confinement ball for the sharp local modulus, and support identification combined with a KL argument yields whole-sequence convergence.

math.OC↗

Weakly Quasar-Convex Optimization

This paper introduces the class of weakly quasar-convex functions, a notion that simultaneously generalizes weak convexity (and hence convexity) and quasar-convexity. We establish the theoretical foundation for this new class by developing its basic calculus, deriving characterizations, and systematically investigating its closure and structural properties. Building on these tools, we analyze the proximal point algorithm for minimizing a weakly quasar-convex function under local quadratic growth, proving $Q$-linear convergence of the iterates and $R$-linear convergence of the function values to a global minimizer, with iteration complexity $O(\ln(\varepsilon^{-1}))$; we also give a fully computable, minimizer-free stopping rule via the proximal residual with matching complexity. Finally, we construct a weakly quasar-convex function, satisfying all our standing assumptions, that is neither weakly convex, quasi-convex, star quasi-convex, nor quasar-convex for any modulus, showing that our convergence theory applies to instances outside every prior framework in the literature.

math.OC↗

Asymptotic Convergence Analysis of High-Order Proximal-Point Methods Beyond Sublinear Rates

This paper investigates the asymptotic convergence behavior of the high-order proximal-point algorithm (HiPPA) to global minimizers, extending existing analyses beyond sublinear convergence rates and complexity analysis. Specifically, we study the proximal operator of a proper lower semicontinuous function augmented with a $p$th-order regularization for $p>1$, and establish the convergence of HiPPA to a global minimizer with a particular focus on its convergence rate. To this end, we focus on minimizing functions in the class of uniformly quasiconvex functions, which includes strongly convex, uniformly convex, and strongly quasiconvex functions as special cases. Our analysis reveals the following convergence behaviors of HiPPA when the uniform quasiconvexity modulus $ϕ$ admits a power function of degree $q$ as a lower bound, i.e., $ϕ(t) \geq c t^q$ for some $c>0$, on an interval $\mathcal{I}$: (i) for $q\in (1,2)$ and $\mathcal{I}=[0,1)$, HiPPA exhibits a local linear rate for $p\in [q,2)$; (ii) HiPPA converges linearly when $p=2$, $q=2$, and also when $p=q>2$, provided that $\mathcal{I}=[0,\infty)$; (iii) for $q\geq 2$ and $\mathcal{I}=[0,\infty)$, HiPPA achieves a superlinear rate for $p>q$. Notably, to our knowledge, some of these results are novel, even in the context of strongly or uniformly convex functions, offering new insights into optimizing generalized convex problems.

math.OC↗

On Debreu-Koopmans Theorem for Additively Decomposed Quasiconvex Functions with Applications

The Debreu Koopmans theorem restricts separable aggregation to at most one nonconvex component. We solve this by proving that a separable, additive or multiplicative, function is star quasiconvex, those with star shaped sublevel sets about minimizers, if and only if each component is star quasiconvex. This immediately yields star quasiconvexity of separable sums of quasiconvex functions, formally bridging diversification theory with the S shaped value functions of Prospect Theory. Furthermore, we develop a complete calculus, monotonic composition, pointwise minima, quasi arithmetic means, and we apply it to Cobb-Douglas functions, multifactor risk models, and constant function market makers in decentralized finance. Star quasiconvexity thus provides a unified framework for applications in optimization and economic modeling beyond the classical Debreu Koopmans constraint. The introduction discuss economic motivations.

math.OC↗

Star Quasiconvexity: a Unified Approach for Linear Convergence of First-Order Methods Beyond Convexity

We introduce and study a new class of generalized convex functions termed star quasiconvex functions. This class includes convex, star-convex, quasiconvex, quasar-convex, and positively homogeneous functions of any degree $p>0$ as special cases. Furthermore, we provide several characterizations of this class covering both nonsmooth and differentiable cases. In particular, in the general nonsmooth case, star quasiconvex functions are characterized by functions for which all its sublevel sets are star-shaped at a minimizer, while in the differentiable case strongly star quasiconvex functions coincides with those satisfying the restricted secant inequality property, and star quasiconvex functions are related with variationally coherent functions, thereby providing a rich framework for differentiable first-order methods. Additionally, we develop standard properties of the proximity operator and prove that the proximal point algorithm converges linearly to the unique solution when applied to strongly star quasiconvex functions defined over closed star-shaped sets that are not necessarily convex.

math.OC↗

Robust Learning Meets Quasar-Convex Optimization: Inexact High-Order Proximal-Point Methods

Robust learning aims to maintain model performance under noise, corruption, and distributional shifts, which are prevalent in modern machine learning applications. This work shows that examples of robust learning problems can be formulated as (strongly) quasar-convex optimization problems, which admit a benign landscape with no saddle points. We then propose HiPPA, an inexact high-order proximal-point method that employs a model-value gap to control the inexactness of subproblem solutions. Notably, we prove global convergence of HiPPA to global minima and establish that it attains a (local) linear or superlinear convergence rate, depending on the regularization order and inexactness control. Our numerical experiments on robust feature-alignment distillation indicate strong empirical performance of HiPPA and results consistent with our theoretical findings.

math.OC↗

Quasar-Convex Optimization: Fundamental Properties and High-Order Proximal-Point Methods

We study the optimization of (strongly) quasar-convex functions, a class that arises naturally in many machine learning and data science applications due to its favorable properties. The fundamental properties of this class are first developed, including its stability under standard calculus operations, growth conditions, and the absence of spurious critical points, which together imply a benign global geometry with no saddle points. Motivated by these properties, a class of proximal-point algorithms (HiPPA) with high-order regularization of order $p>1$ is introduced. Conditions are identified under which the iterates converge globally to minimizers, and a unified convergence analysis is provided with explicit rates and iteration complexity bounds under appropriate regularity assumptions. The results reveal a sharp transition in behavior with respect to the order $p$: for $p\in(1,2)$, the method achieves local linear convergence with complexity $\mathcal{O}(\log(\varepsilon^{-1}))$ when initialized sufficiently close to a minimizer; for $p=2$, it attains global linear convergence with the same complexity; and for $p>2$, it exhibits superlinear convergence with complexity $\mathcal{O}(\log\log(\varepsilon^{-1}))$, where $\varepsilon>0$ denotes the target accuracy. The theory is complemented with preliminary numerical experiments on selected machine learning problems, which illustrate the effectiveness of the proposed methods and are consistent with the theoretical findings.

math.OC↗

Extending Linear Convergence of the Proximal Point Algorithm: The Quasar-Convex Case

This work investigates the properties of the proximity operator for quasar-convex functions and establishes the convergence of the proximal point algorithm to a global minimizer with a particular focus on its convergence rate. In particular, we demonstrate: (i) the generated sequence is mi\-ni\-mi\-zing and achieves an $\mathcal{O}(\varepsilon^{-1})$ complexity rate for quasar-convex functions; (ii) under strong quasar-convexity, the sequence converges linearly and attains an $\mathcal{O}(\ln(\varepsilon^{-1}))$ complexity rate. These results extend known convergence rates from the (strongly) convex to the (strongly) quasar-convex setting. To the best of our knowledge, some findings are novel even for the special case of (strongly) star-convex functions. Numerical experiments corroborate our theoretical results.

math.OC↗

On Optimality Conditions for Mathematical Programming Problems Based on Strong Subdifferentials

We develop refined Karush-Kuhn-Tucker (KKT) and Fritz-John (FJ)-type optimality conditions for nonsmooth, nonconvex mathematical pro\-gra\-mming problems. We pay special attention in the case that the functional constraint belongs to a specific class of generalized convex functions known as strongly quasiconvex functions. After analyzing a specialized sub\-di\-ffe\-ren\-tial, named the strong subdifferential, we compute the normal cone of the supremum function in terms of such subdifferentials, and apply this result to the mathematical programming problem. We illustrate our important results by examples.

math.OC↗

Optimal Routing across Constant Function Market Makers with Gas Fees

We study the optimal routing problem in decentralized exchanges built on Constant Function Market Makers when trades can be split across multiple heterogeneous pools and execution incurs fixed on-chain costs (gas fees). While prior routing formulations typically abstract from fixed activation costs, real on-chain execution presents non-negligible gas fees. They also become convex under concavity/convexity assumptions on the invariant functions. We propose a general optimization framework that allows differentiable invariant functions beyond global convexity and incorporates fixed gas fees through a mixed-integer model that induces activation thresholds. Subsequently, we introduce a relaxed formulation of this model, whereby we deduce necessary optimality conditions, obtaining an explicit Karush-Kuhn-Tucker system that links prices, fees, and activation. We further establish sufficient optimality conditions using tools from generalized convexity (pseudoconcavity/pseudoconvexity and quasilinearity), yielding a verifiable optimality characterization without requiring convex trade functions. Finally, we relate the relaxed solution to the original mixed-integer model by providing explicit approximation bounds that quantify the utility gap induced by relaxation. Our results extend the mathematical theory for routing by offering no-trade conditions in fragmented on-chain markets in the presence of gas fees.

math.OC↗

Splitting Proximal Point Algorithms for the Sum of Prox-Convex Functions

This paper addresses the minimization of a finite sum of prox-convex functions under Lipschitz continuity of each component. We propose two variants of the splitting proximal point algorithms proposed in \cite{Bacak,Bertsekas}: one deterministic with a fixed update order, and one stochastic with random sampling, and we extend them from convex to prox-convex functions. We prove global convergence for both methods under standard stepsize a\-ssump\-tions, with almost sure convergence for the stochastic variant via supermartingale theory. Numerical experiments with nonconvex quadratic functions illustrate the efficiency of the proposed methods and support the theoretical results.

math.OC↗

Delayed Feedback in Online Non-Convex Optimization: A Non-Stationary Approach with Applications

We study non-convex delayed-noise online optimization problems by evaluating dynamic regret in the non-stationary setting when the loss functions are quasar-convex. In particular, we consider scenarios involving quasar-convex functions either with a Lipschitz gradient or weakly smooth and, for each case, we ensure bounded dynamic regret in terms of cumulative path variation achieving sub-linear regret rates. Furthermore, we illustrate the flexibility of our framework by applying it to both theoretical settings such as zeroth-order (bandit) and also to practical applications with quadratic fractional functions. Moreover, we provide new examples of non-convex functions that are quasar-convex by proving that the class of differentiable strongly quasiconvex functions (Polyak 1966) are strongly quasar-convex on convex compact sets. Finally, several numerical experiments validate our theoretical findings, illustrating the effectiveness of our approach.

math.OC↗

Discontinuous Strongly Quasiconvex Functions

A fundamental open question asking whether all real-valued strongly quasiconvex functions defined on $\mathbb R^n$ are necessarily continuous, akin to their convex counterparts, is answered in detail in this paper. Among other things, we show that such functions can have infinitely many points of discontinuity. The failure of lower semicontinuity together with the lack of upper semicontinuity at infinitely many points of certain real-valued strongly quasiconvex functions are also shown.

math.OC↗

Optimality Conditions and Duality for Multiobjective Fractional Bilevel Optimization Problems

This paper studies a multiobjective bilevel optimization problem where each objective is a fractional function. By reformulating the problem into a single-level one, we establish refined necessary and sufficient optimality conditions. These results are derived using ${\partial}_D$-nonsmooth Abadie-type constraint qualifications and generalized convexity concepts (quasiconvexity and pseudoconvexity) based on directional convexificators. We also prove weak and strong duality theorems for a Mond-Weir dual problem formulated with directional convexificators. Finally, several examples are provided to illustrate the advantages of our approach.

math.OC↗

Characterizations of Strongly Quasiconvex Functions

We provide new necessary and sufficient conditons for ensuring strong quasiconvexity in the nonsmooth case and, as a consequence, we provide a proof for the differentiable case. Furthermore, we improve the quadratic growth property for strongly quasiconvex functions.

math.OC↗

Regular Subgradients of Marginal Functions with Applications to Calculus and Bilevel Programming

The paper addresses the study and applications of a broad class of extended-real-valued functions, known as optimal value or marginal functions, which are frequently appeared in variational analysis, parametric optimization, and a variety of applications. Functions of this type are intrinsically nonsmooth and require the usage of tools of generalized differentiation. The main results of this paper provide novel evaluations and exact calculations of regular/Fréchet subgradients and their singular counterparts for general classes of marginal functions via their given data. The obtained results are applied to establishing new calculus rules for such subgradients and necessary optimality conditions in bilevel programming

math.OC↗