-Lusin approximation of strongly convex functions

Daniel Azagra,Marjorie Drake,Piotr Hajłasz

DOI: https://doi.org/10.1007/s00222-024-01252-6

IF: 3.128

2024-04-04

Inventiones mathematicae

Abstract:We prove that if is strongly convex, then for every there is a strongly convex function such that and .

mathematics

What problem does this paper attempt to address?

$\mathbf{C^2}$-Lusin approximation of strongly convex functions

Daniel Azagra,Marjorie Drake,Piotr Hajłasz

2024-03-25

Abstract:We prove that if $u:\mathbb{R}^n\to\mathbb{R}$ is strongly convex, then for every $\varepsilon>0$ there is a strongly convex function $v\in C^2(\mathbb{R}^n)$ such that $|\{u\neq v\}|<\varepsilon$ and $\Vert u-v\Vert_\infty<\varepsilon$.

Classical Analysis and ODEs
$\mathbf{C^2}$-Lusin approximation of strongly convex bodies

Daniel Azagra,Marjorie Drake,Piotr Hajłasz

2024-10-11

Abstract:We prove that, if $W \subset \mathbb{R}^n$ is a locally strongly convex body (not necessarily compact), then for any open set $V \supset \partial W$ and $\varepsilon>0$, and $V \supset \partial W$ is open, then there exists a $C^2$ locally strongly convex body $W_{\varepsilon, V}$ such that $\mathcal{H}^{n-1}(\partial W_{\varepsilon, V}\triangle\,\partial W)<\varepsilon$ and $\partial W_{\varepsilon, V}\subset V$. Moreover, if $W$ is strongly convex, then $W_{\varepsilon, V}$ is strongly convex as well.

Classical Analysis and ODEs,Differential Geometry
Strongly quasiconvex functions: what we know (so far)

Sorin-Mihai Grad,Felipe Lara,Raúl T. Marcavillaca

2024-10-30

Abstract:Introduced by Polyak in 1966, the class of strongly quasiconvex functions includes some interesting nonconvex members, like the square root of the Euclidean norm or ratios with a nonnegative strongly convex numerator and a concave and positive denominator. This survey collects the vast majority of the results involving strongly quasiconvex functions available in the literature at the moment, presenting, in particular, algorithms for minimizing such functions, and suggests some directions where additional investigations would be welcome.

Optimization and Control
A geometric approach to second-order differentiability of convex functions

Daniel Azagra,Anthony Cappello,Piotr Hajłasz

2023-08-01

Abstract:We show a new, elementary and geometric proof of the classical Alexandrov theorem about the second order differentiability of convex functions. We also show new proofs of recent results about Lusin approximation of convex functions and convex bodies by $C^{1,1}$ convex functions and convex bodies.

Classical Analysis and ODEs,Differential Geometry
NONLINEAR SIMULTANEOUS APPROXIMATION IN COMPLETE LATTICE BANACH SPACES

Donghui Fang,Xianfa Luo,Chong Li

DOI: https://doi.org/10.11650/twjm/1500405185

2008-01-01

Taiwanese Journal of Mathematics

Abstract:This paper is concerned with the problem of nonlinear best simultaneous approximations in conditional complete lattice Banach spaces with a strong unit. Characterization results of the best simultaneous approximation from simultaneous suns and suns are established. A counterexample, to which the characterization theorem for convex sets due to Mohebi (Numer. Funct. Anal. Optim., 25 (2004), 685-705) fails, is provided and a corrected version of the theorem is presented.
Stable and Total Fenchel Duality for Convex Optimization Problems in Locally Convex Spaces

Chong Li,Donghui Fang,Genaro Lopez,Marco A. Lopez

DOI: https://doi.org/10.1137/080734352

IF: 2.763

2009-01-01

SIAM Journal on Optimization

Abstract:We consider the optimization problem $(P_{A})$ $\inf_{x\in X}\{f(x)+g(Ax)\}$ where $f$ and $g$ are proper convex functions defined on locally convex Hausdorff topological vector spaces $X$ and $Y$, respectively, and $A$ is a linear operator from $X$ to $Y$. By using the properties of the epigraph of the conjugated functions, some sufficient and necessary conditions for the strong Fenchel duality and the strong converse Fenchel duality of $(P_{A})$ are provided. Sufficient and necessary conditions for the stable Fenchel duality and for the total Fenchel duality are also derived.
Lusin's Theorem and Bochner Integration

Peter A. Loeb,Erik Talvila

DOI: https://doi.org/10.48550/arXiv.math/0406370

2004-06-18

Classical Analysis and ODEs

Abstract:It is shown that the approximating functions used to define the Bochner integral can be formed using geometrically nice sets, such as balls, from a differentiation basis. Moreover, every appropriate sum of this form will be within a preassigned $\epsilon$ of the integral, with the sum for the local errors also less than $\epsilon$. All of this follows from the ubiquity of Lebesgue points, which is a consequence of Lusin's theorem, for which a simple proof is included in the discussion.
Strong Uniqueness of Best Approximations in Spaces of Bounded Linear Operators

Li Chong,He Jinsu

DOI: https://doi.org/10.1360/02ys0336

2004-01-01

Science in China Series A Mathematics

Abstract:The present paper is concerned with problems of the strong uniqueness of the best approximation and the characterization of a uniqueness element in operator spaces. Some results on the strong uniqueness of the best approximation operator from RS-sets are proved and the uniqueness element of a sun in the compact operator space from c(0) to c(0) is characterized by the strict Kolmogorov's condition. Some recent results due to Lewicki and others are extended and improved.
The Equivalence of Three Types of Error Bounds for Weakly and Approximately Convex Functions

Bai, Sixuan,Li, Minghua,Lu, Chengwu,Zhu, Daoli,Deng, Sien

DOI: https://doi.org/10.1007/s10957-022-02016-z

2022-01-01

Journal of Optimization Theory and Applications

Abstract:We start by establishing the equivalence of three types of error bounds: weak sharp minima, level-set subdifferential error bounds and Łojasiewicz (for short Ł) inequalities for weakly convex functions with exponent α ∈ [ 0 , 1 ] and approximately convex functions. Then we apply these equivalence results to a class of nonconvex optimization problems, whose objective functions are the sum of a convex function and a composite function with a locally Lipschitz function and a smooth vector-valued function. Finally, applying a characterization for lower-order regularization problems, we show that the level-set subdifferential error bound with exponent 1 and the Ł inequality with exponent 1 2 hold at a local minimum point.
Constraint Qualifications for Extended Farkas'S Lemmas and Lagrangian Dualities in Convex Infinite Programming

D. H. Fang,C. Li,K. F. Ng

DOI: https://doi.org/10.1137/080739124

IF: 2.763

2010-01-01

SIAM Journal on Optimization

Abstract:For an inequality system defined by a possibly infinite family of proper functions (not necessarily lower semicontinuous), we introduce some new notions of constraint qualifications in terms of the epigraphs of the conjugates of these functions. Under the new constraint qualifications, we obtain characterizations of those reverse-convex inequalities which are a consequence of the constrained system, and we provide necessary and/or sufficient conditions for a stable Farkas lemma to hold. Similarly, we provide characterizations for constrained minimization problems to have the strong or strong stable Lagrangian dualities. Several known results in the conic programming problem are extended and improved.
A Fenchel-Moreau Theorem for (L)over-bar(0)-Valued Functions

Samuel Drapeau,Asgar Jamneshan,Michael Kupper

2019-01-01

Journal of convex analysis

Abstract:We establish a Fenchel-Moreau type theorem for proper convex functions f : X -> (L) over bar (0), where (X,Y,<.,.>) is a dual pair of Banach spaces and (L) over bar (0) is the space of all extended real-valued functions on a sigma-finite measure space. We introduce the concept of stable lower semi-continuity which is shown to be equivalent to the existence of a dual representation f(x) = sup(y is an element of L0(Y)) {< x, y > - f*(y)}, x is an element of X, where L-0(Y) is the space of all strongly measurable functions with values in Y, and <.,.> is understood pointwise almost everywhere. The proof is based on a conditional extension result and conditional functional analysis.
Finiteness Principles for Smooth Convex Functions

Marjorie K. Drake

2024-02-26

Abstract:Let $E \subset \mathbb{R}^n$ be a compact set, and $f:E \to \mathbb{R}$. How can we tell if there exists a convex extension $F \in C^{1,1}(\mathbb{R}^n)$ of $f$, i.e. satisfying $F|_E = f|_E$? Assuming such an extension exists, how small can one take the Lipschitz constant $\text{Lip}(\nabla F): = \sup_{x,y \in \mathbb{R}^n, x \neq y} \frac{|\nabla F(x) - \nabla F(y)|}{|x-y|}$? We provide an answer to these questions for the class of strongly convex functions by proving that there exist constants $k^\# \in \mathbb{N}$ and $C>0$ depending only on the dimension $n$, such that if for every subset $S \subset E$, $\#S \leq k^\#$, there exists an $\eta$-strongly convex function $F^S \in C^{1,1}(\mathbb{R}^n)$ satisfying $F^S|_S=f|_S$ and $\text{Lip}(\nabla F^S) \leq M$, then there exists an ${\frac{\eta}{C}}$-strongly convex function $F \in C^{1,1}_c(\mathbb{R}^n)$ satisfying $F|_E = f|_E$, and $\text{Lip}(\nabla F) \leq C M^2/\eta$. Further, we prove a Finiteness Principle for the space of convex functions in $C^{1,1}(\mathbb{R})$ and that the sharp finiteness constant for this space is $k^\#=5$.

Classical Analysis and ODEs
Stochastic Approximation of Smooth and Strongly Convex Functions: Beyond the Convergence Rate

Lijun Zhang,Zhi-Hua Zhou

2019-01-01

Abstract:Stochastic approximation (SA) is a classical approach for stochastic convex optimization. Previous studies have demonstrated that the convergence rate of SA can be improved by introducing either smoothness or strong convexity condition. In this paper, we make use of smoothness and strong convexity simultaneously to boost the convergence rate. Let be the modulus of strong convexity, be the condition number, be the minimal risk, and be some small constant. First, we demonstrate that, in expectation, an risk bound is attainable when . Thus, when is small, the convergence rate could be faster than and approaches in the ideal case. Second, to further benefit from small risk, we show that, in expectation, an risk bound is achievable. Thus, the excess risk reduces exponentially until reaching , and if , we obtain a global linear convergence. Finally, we emphasize that our proof is constructive and each risk bound is equipped with an efficient stochastic algorithm attaining that bound.
A Fenchel-Moreau theorem for $\bar L^0$-valued functions

Samuel Drapeau,Asgar Jamneshan,Michael Kupper

2017-01-01

Abstract: We establish a Fenchel-Moreau theorem for proper convex functions $f\colon X\to \bar{L}^0$, where $(X,Y,\langle \cdot,\cdot \rangle)$ is a dual pair of Banach spaces and $\bar L^0$ is the space of all extended real-valued functions on a $\sigma$-finite measure space. We introduce the concept of stable lower semi-continuity which is shown to be equivalent to the existence of a dual representation \[ f(x)=\sup_{y \in L^0(Y)} \left\{\langle x, y \rangle - f^\ast(y)\right\}, \quad x\in X, \] where $L^0(Y)$ is the space of all strongly measurable functions with values in $Y$, and $\langle \cdot,\cdot \rangle$ is understood pointwise almost everywhere.
The Minimizer of the Sum of Two Strongly Convex Functions

Kananart Kuwaranancharoen,Shreyas Sundaram

DOI: https://doi.org/10.1080/02331934.2024.2402923

2024-09-22

Abstract:The optimization problem concerning the determination of the minimizer for the sum of convex functions holds significant importance in the realm of distributed and decentralized optimization. In scenarios where full knowledge of the functions is not available, limiting information to individual minimizers and convexity parameters -- either due to privacy concerns or the nature of solution analysis -- necessitates an exploration of the region encompassing potential minimizers based solely on these known quantities. The characterization of this region becomes notably intricate when dealing with multivariate strongly convex functions compared to the univariate case. This paper contributes outer and inner approximations for the region harboring the minimizer of the sum of two strongly convex functions, given a constraint on the norm of the gradient at the minimizer of the sum. Notably, we explicitly delineate the boundaries and interiors of both the outer and inner approximations. Intriguingly, the boundaries as well as the interiors turn out to be identical. Furthermore, we establish that the boundary of the region containing potential minimizers aligns with that of the outer and inner approximations.

Optimization and Control
Lusin spaces as images of locally compact Polish spaces

Alina Kargol,Yuri Kozitsky

2023-07-31

Abstract:A Lusin space is a Hausdorff space being the image of a Polish space under a continuous bijection. Such spaces have multiple applications, in particular, as state spaces of various stochastic systems. In this work, we consider the spaces obtained as the images of a noncompact and locally compact Polish space $(X, \mathcal{T})$, which we call $c$-Lusin. The main result is the statement that a $c$-Lusin space $Y=f(X)$, can be written as $Z\cup Y_1$, where $Z$ is a locally compact Polish space whereas $Y_1$ is $c$-Lusin. At the same time, $Y_1$ is the set of the discontinuity points of $f^{-1}$ which is a closed subset of $Y$. Moreover, $Y_1$ is nowhere dense if (and only if) $Y$ is a Baire space. By the same arguments, $Y_1$ can also be decomposed as $Z_1 \cup Y_2$ with the properties as above. In the case where $f$ can be extended to a continuous map $f:X\cup \{\infty\} \to Y$, and thus $Y_1$ is a singleton, we construct a metric on $X$ such that the corresponding metric space is compact and homeomorphic to the $c$-Lusin space $(f(X), \mathcal{T}')$.

General Topology
Order preserving and order reversing operators on the class of $L^0$-convex functions in complete random normed modules

Mingzhi Wu,Tiexin Guo,Long Long

DOI: https://doi.org/10.48550/arXiv.2206.06071

2022-06-13

Abstract:Based on both the fundamental theorem of affine geometry in regular $L^0$-modules and the recent progress in random convex analysis, this paper characterizes the stable fully order preserving and order reversing operators acting on the class of proper lower semicontinuous $L^0$-convex functions in complete random normed modules.

Functional Analysis
Error bounds, PL condition, and quadratic growth for weakly convex functions, and linear convergences of proximal point methods

Feng-Yi Liao,Lijun Ding,Yang Zheng

2024-08-14

Abstract:Many practical optimization problems lack strong convexity. Fortunately, recent studies have revealed that first-order algorithms also enjoy linear convergences under various weaker regularity conditions. While the relationship among different conditions for convex and smooth functions is well-understood, it is not the case for the nonsmooth setting. In this paper, we go beyond convexity and smoothness, and clarify the connections among common regularity conditions in the class of weakly convex functions, including $\textit{strong convexity}$, $\textit{restricted secant inequality}$, $\textit{subdifferential error bound}$, $\textit{Polyak-Łojasiewicz inequality}$, and $\textit{quadratic growth}$. In addition, using these regularity conditions, we present a simple and modular proof for the linear convergence of the proximal point method (PPM) for convex and weakly convex optimization problems. The linear convergence also holds when the subproblems of PPM are solved inexactly with a proper control of inexactness.

Optimization and Control,Systems and Control
On affine Lusztig varieties

Xuhua He

2024-10-08

Abstract:Affine Lusztig varieties encode the orbital integrals of Iwahori--Hecke functions and serve as building blocks for the (conjectural) theory of affine character sheaves. We establish a close relationship between affine Lusztig varieties and affine Deligne--Lusztig varieties. Consequently, we give an explicit nonemptiness pattern and dimension formula for affine Lusztig varieties in most cases.

Representation Theory,Algebraic Geometry,Number Theory

-Lusin approximation of strongly convex functions

$\mathbf{C^2}$-Lusin approximation of strongly convex functions

$\mathbf{C^2}$-Lusin approximation of strongly convex bodies

Strongly quasiconvex functions: what we know (so far)

A geometric approach to second-order differentiability of convex functions

NONLINEAR SIMULTANEOUS APPROXIMATION IN COMPLETE LATTICE BANACH SPACES

Stable and Total Fenchel Duality for Convex Optimization Problems in Locally Convex Spaces

Lusin's Theorem and Bochner Integration

Strong Uniqueness of Best Approximations in Spaces of Bounded Linear Operators

The Equivalence of Three Types of Error Bounds for Weakly and Approximately Convex Functions

Constraint Qualifications for Extended Farkas'S Lemmas and Lagrangian Dualities in Convex Infinite Programming

A Fenchel-Moreau Theorem for (L)over-bar(0)-Valued Functions

Finiteness Principles for Smooth Convex Functions

Stochastic Approximation of Smooth and Strongly Convex Functions: Beyond the Convergence Rate

A Fenchel-Moreau theorem for $\bar L^0$-valued functions

The Minimizer of the Sum of Two Strongly Convex Functions

Lusin spaces as images of locally compact Polish spaces

Order preserving and order reversing operators on the class of $L^0$-convex functions in complete random normed modules

Error bounds, PL condition, and quadratic growth for weakly convex functions, and linear convergences of proximal point methods

On affine Lusztig varieties