Litcius/Paper detail

An inexact regularized proximal Newton method for nonconvex and nonsmooth optimization

Ruyu Liu, Shaohua Pan, Yuqia Wu, Xiaoqi Yang

2024Computational Optimization and Applications13 citationsDOIOpen Access PDF

Abstract

Abstract This paper focuses on the minimization of a sum of a twice continuously differentiable function f and a nonsmooth convex function. An inexact regularized proximal Newton method is proposed by an approximation to the Hessian of f involving the $$\varrho $$ <mml:math xmlns:mml="http://www.w3.org/1998/Math/MathML"> <mml:mi>ϱ</mml:mi> </mml:math> th power of the KKT residual. For $$\varrho =0$$ <mml:math xmlns:mml="http://www.w3.org/1998/Math/MathML"> <mml:mrow> <mml:mi>ϱ</mml:mi> <mml:mo>=</mml:mo> <mml:mn>0</mml:mn> </mml:mrow> </mml:math> , we justify the global convergence of the iterate sequence for the KL objective function and its R-linear convergence rate for the KL objective function of exponent 1/2. For $$\varrho \in (0,1)$$ <mml:math xmlns:mml="http://www.w3.org/1998/Math/MathML"> <mml:mrow> <mml:mi>ϱ</mml:mi> <mml:mo>∈</mml:mo> <mml:mo>(</mml:mo> <mml:mn>0</mml:mn> <mml:mo>,</mml:mo> <mml:mn>1</mml:mn> <mml:mo>)</mml:mo> </mml:mrow> </mml:math> , by assuming that cluster points satisfy a locally Hölderian error bound of order q on a second-order stationary point set and a local error bound of order $$q&gt;1\!+\!\varrho $$ <mml:math xmlns:mml="http://www.w3.org/1998/Math/MathML"> <mml:mrow> <mml:mi>q</mml:mi> <mml:mo>&gt;</mml:mo> <mml:mn>1</mml:mn> <mml:mspace/> <mml:mo>+</mml:mo> <mml:mspace/> <mml:mi>ϱ</mml:mi> </mml:mrow> </mml:math> on the common stationary point set, respectively, we establish the global convergence of the iterate sequence and its superlinear convergence rate with order depending on q and $$\varrho $$ <mml:math xmlns:mml="http://www.w3.org/1998/Math/MathML"> <mml:mi>ϱ</mml:mi> </mml:math> . A dual semismooth Newton augmented Lagrangian method is also developed for seeking an inexact minimizer of subproblems. Numerical comparisons with two state-of-the-art methods on $$\ell _1$$ <mml:math xmlns:mml="http://www.w3.org/1998/Math/MathML"> <mml:msub> <mml:mi>ℓ</mml:mi> <mml:mn>1</mml:mn> </mml:msub> </mml:math> -regularized Student’s t -regressions, group penalized Student’s t -regressions, and nonconvex image restoration confirm the efficiency of the proposed method.

Topics & Concepts

MathematicsMathematical optimizationNewton's methodApplied mathematicsNonlinear systemPhysicsQuantum mechanicsSparse and Compressive Sensing TechniquesAdvanced Optimization Algorithms ResearchNumerical methods in inverse problems