Linear Convergence Rate of Class of Distributed Augmented Lagrangian Algorithms

نویسنده

  • Dušan Jakovetić
چکیده

We study distributed optimization where nodes cooperatively minimize the sum of their individual, locally known, convex costs fi(x)’s, x ∈ R is global. Distributed augmented Lagrangian (AL) methods have good empirical performance on several signal processing and learning applications, but there is limited understanding of their convergence rates and how it depends on the underlying network. This paper establishes globally linear (geometric) convergence rates of a class of deterministic and randomized distributed AL methods, when the fi’s are twice continuously differentiable and have a bounded Hessian. We give explicit dependence of the convergence rates on the underlying network parameters. Simulations illustrate our analytical findings.

برای دانلود رایگان متن کامل این مقاله و بیش از 32 میلیون مقاله دیگر ابتدا ثبت نام کنید

ثبت نام

اگر عضو سایت هستید لطفا وارد حساب کاربری خود شوید

منابع مشابه

Augmented Lagrangian method for solving absolute value equation and its application in two-point boundary value problems

One of the most important topic that consider in recent years by researcher is absolute value equation (AVE). The absolute value equation seems to be a useful tool in optimization since it subsumes the linear complementarity problem and thus also linear programming and convex quadratic programming. This paper introduce a new method for solving absolute value equation. To do this, we transform a...

متن کامل

THE EFFECTS OF INITIAL SAMPLING AND PENALTY FUNCTIONS IN OPTIMAL DESIGN OF TRUSSES USING METAHEURISTIC ALGORITHMS

Although Genetic algorithm (GA), Ant colony (AC) and Particle swarm optimization algorithm (PSO) have already been extended to various types of engineering problems, the effects of initial sampling beside constraints in the efficiency of algorithms, is still an interesting field. In this paper we show that, initial sampling with a special series of constraints play an important role in the conv...

متن کامل

An Augmented Lagrangian Based Algorithm for Distributed NonConvex Optimization

This paper is about distributed derivative-based algorithms for solving optimization problems with a separable (potentially nonconvex) objective function and coupled affine constraints. A parallelizable method is proposed that combines ideas from the fields of sequential quadratic programming and augmented Lagrangian algorithms. The method negotiates shared dual variables that may be interprete...

متن کامل

Local Convergence Properties of

We consider the local convergence properties of the class of augmented Lagrangian methods for solving nonlinear programming problems whose global convergence properties are analyzed by Conn et al. (1993a). In these methods, linear constraints are treated separately from more general constraints. These latter constraints are combined with the objective function in an augmented Lagrangian while t...

متن کامل

Local Convergence Properties of two Augmented Lagrangian Algorithms for Optimization with a Combination of General Equality and Linear Constraints

We consider the local convergence properties of the class of augmented Lagrangian methods for solving nonlinear programming problems whose global convergence properties are analyzed by Conn et al. (1993a). In these methods, linear constraints are treated separately from more general constraints. These latter constraints are combined with the objective function in an augmented Lagrangian while t...

متن کامل

ذخیره در منابع من


  با ذخیره ی این منبع در منابع من، دسترسی به آن را برای استفاده های بعدی آسان تر کنید

عنوان ژورنال:

دوره   شماره 

صفحات  -

تاریخ انتشار 2014