arXiv subjects

Ion Necoara

Publications and source records attributed to Ion Necoara.

At least 55 records · Page 3Linked to original sources

arXiv2019

Linear convergence of dual coordinate descent on non-polyhedral convex problems

This paper deals with constrained convex problems, where the objective function is smooth strongly convex and the feasible set is given as the intersection of a large number of closed convex (possibly non-polyhedral) sets. In order to deal efficiently with the complicated constraints we consider a dual formulation of this problem. We prove that the corresponding dual function satisfies a quadratic growth property on any sublevel set, provided that the objective function is smooth and strongly convex and the sets verify the Slater's condition. To the best of our knowledge, this work is the first deriving a quadratic growth condition for the dual under these general assumptions. Existing works derive similar quadratic growth conditions under more conservative assumptions, e.g., the sets need to be either polyhedral or compact. Then, for finding the minimum of the dual problem, due to its special composite structure, we propose random (accelerated) coordinate descent algorithms. However, with the existing theory one can prove that such methods converge only sublinearly. Based on our new quadratic growth property derived for the dual, we now show that such methods have faster convergence, that is the dual random (accelerated) coordinate descent algorithms converge linearly. Besides providing a general dual framework for the analysis of randomized coordinate descent schemes, our results resolve an open problem in the literature related to the convergence of Dykstra algorithm on the best feasibility problem for a collection of convex sets. That is, we establish linear convergence rate for the randomized Dykstra algorithm when the convex sets satisfy the Slater's condition and derive also a new accelerated variant for the Dykstra algorithm.

math.OC↗

arXiv2019

Minibatch stochastic subgradient-based projection algorithms for solving convex inequalities

This paper deals with the convex feasibility problem, where the feasible set is given as the intersection of a (possibly infinite) number of closed convex sets. We assume that each set is specified algebraically as a convex inequality, where the associated convex function is general (possibly non-differentiable). For finding a point satisfying all the convex inequalities we design and analyze random projection algorithms using special subgradient iterations and extrapolated stepsizes. Moreover, the iterate updates are performed based on parallel random observations of several constraint components. For these minibatch stochastic subgradient-based projection methods we prove sublinear convergence results and, under some linear regularity condition for the functional constraints, we prove linear convergence rates. We also derive conditions under which these rates depend explicitly on the minibatch size. To the best of our knowledge, this work is the first deriving conditions that show when minibatch stochastic subgradient-based projection updates have a better complexity than their single-sample variants.

math.OC↗

arXiv2019

Faster randomized block Kaczmarz algorithms

The Kaczmarz algorithm is a simple iterative scheme for solving consistent linear systems. At each step, the method projects the current iterate onto the solution space of a single constraint. Hence, it requires very low cost per iteration and storage, and it has a linear rate of convergence. Distributed implementations of Kaczmarz have become, in recent years, the de facto architectural choice for large-scale linear systems. Therefore, in this paper we develop a family of randomized block Kaczmarz algorithms that uses at each step a subset of the constraints and extrapolated stepsizes, and can be deployed on distributed computing units. Our approach is based on several new ideas and tools, including stochastic selection rule for the blocks of rows, stochastic conditioning of the linear system, and novel strategies for designing extrapolated stepsizes. We prove that randomized block Kaczmarz algorithm converges linearly in expectation, with a rate depending on the geometric properties of the matrix and its submatrices and on the size of the blocks. Our convergence analysis reveals that the algorithm is most effective when it is given a good sampling of the rows into well-conditioned blocks. Besides providing a general framework for the design and analysis of randomized block Kaczmarz methods, our results resolve an open problem in the literature related to the theoretical understanding of observed practical efficiency of extrapolated block Kaczmarz methods.

math.OC↗

arXiv2019

Almost surely constrained convex optimization

We propose a stochastic gradient framework for solving stochastic composite convex optimization problems with (possibly) infinite number of linear inclusion constraints that need to be satisfied almost surely. We use smoothing and homotopy techniques to handle constraints without the need for matrix-valued projections. We show for our stochastic gradient algorithm $\mathcal{O}(\log(k)/\sqrt{k})$ convergence rate for general convex objectives and $\mathcal{O}(\log(k)/k)$ convergence rate for restricted strongly convex objectives. These rates are known to be optimal up to logarithmic factors, even without constraints. We demonstrate the performance of our algorithm with numerical experiments on basis pursuit, a hard margin support vector machines and a portfolio optimization and show that our algorithm achieves state-of-the-art practical performance.

math.OC↗

arXiv2018

Randomized sketch descent methods for non-separable linearly constrained optimization

In this paper we consider large-scale smooth optimization problems with multiple linear coupled constraints. Due to the non-separability of the constraints, arbitrary random sketching would not be guaranteed to work. Thus, we first investigate necessary and sufficient conditions for the sketch sampling to have well-defined algorithms. Based on these sampling conditions we developed new sketch descent methods for solving general smooth linearly constrained problems, in particular, random sketch descent and accelerated random sketch descent methods. From our knowledge, this is the first convergence analysis of random sketch descent algorithms for optimization problems with multiple non-separable linear constraints. For the general case, when the objective function is smooth and non-convex, we prove for the non-accelerated variant sublinear rate in expectation for an appropriate optimality measure. In the smooth convex case, we derive for both algorithms, non-accelerated and accelerated random sketch descent, sublinear convergence rates in the expected values of the objective function. Additionally, if the objective function satisfies a strong convexity type condition, both algorithms converge linearly in expectation. In special cases, where complexity bounds are known for some particular sketching algorithms, such as coordinate descent methods for optimization problems with a single linear coupled constraint, our theory recovers the best-known bounds. We also show that when random sketch is sketching the coordinate directions randomly produces better results than the fixed selection rule. Finally, we present some numerical examples to illustrate the performances of our new algorithms.

math.OC↗

arXiv2018

Randomized projection methods for convex feasibility problems: conditioning and convergence rates

Finding a point in the intersection of a collection of closed convex sets, that is the convex feasibility problem, represents the main modeling strategy for many computational problems. In this paper we analyze new stochastic reformulations of the convex feasibility problem in order to facilitate the development of new algorithmic schemes. We also analyze the conditioning problem parameters using certain (linear) regularity assumptions on the individual convex sets. Then, we introduce a general random projection algorithmic framework, which extends to the random settings many existing projection schemes, designed for the general convex feasibility problem. Our general random projection algorithm allows to project simultaneously on several sets, thus providing great flexibility in matching the implementation of the algorithm on the parallel architecture at hand. Based on the conditioning parameters, besides the asymptotic convergence results, we also derive explicit sublinear and linear convergence rates for this general algorithmic framework.

math.OC↗

arXiv2017

Nonasymptotic convergence of stochastic proximal point algorithms for constrained convex optimization

A very popular approach for solving stochastic optimization problems is the stochastic gradient descent method (SGD). Although the SGD iteration is computationally cheap and the practical performance of this method may be satisfactory under certain circumstances, there is recent evidence of its convergence difficulties and instability for unappropriate parameters choice. To avoid these drawbacks naturally introduced by the SGD scheme, the stochastic proximal point algorithms have been recently considered in the literature. We introduce a new variant of the stochastic proximal point method (SPP) for solving stochastic convex optimization problems subject to (in)finite intersection of constraints satisfying a linear regularity type condition. For the newly introduced SPP scheme we prove new nonasymptotic convergence results. In particular, for convex and Lipschitz continuous objective functions, we prove nonasymptotic estimates for the rate of convergence in terms of the expected value function gap of order $\mathcal{O}(1/k^{1/2})$, where $k$ is the iteration counter. We also derive better nonasymptotic bounds for the rate of convergence in terms of expected quadratic distance from the iterates to the optimal solution for smooth strongly convex objective functions, which in the best case is of order $\mathcal{O}(1/k)$. Since these convergence rates can be attained by our SPP algorithm only under some natural restrictions on the stepsize, we also introduce a restarting variant of SPP method that overcomes these difficulties and derive the corresponding nonasymptotic convergence rates. Numerical evidence supports the effectiveness of our methods in real-world problems.

math.OC↗

arXiv2017

Complexity of first order inexact Lagrangian and penalty methods for conic convex programming

In this paper we present a complete iteration complexity analysis of inexact first order Lagrangian and penalty methods for solving cone constrained convex problems that have or may not have optimal Lagrange multipliers that close the duality gap. We first assume the existence of optimal Lagrange multipliers and study primal-dual first order methods based on inexact information and augmented Lagrangian smoothing or Nesterov type smoothing. For inexact (fast) gradient augmented Lagrangian methods we derive a total computational complexity of $\mathcal{O}\left(\frac{1}ε\right)$ projections onto a simple primal set in order to attain an $ε-$optimal solution of the conic convex problem. For the inexact fast gradient method combined with Nesterov type smoothing we derive computational complexity $\mathcal{O}\left(\frac{1}{ε^{3/2}}\right)$ projections onto the same set. Then, we assume that optimal Lagrange multipliers for the cone constrained convex problem might not exist, and analyze the fast gradient method for solving penalty reformulations of the problem. For the fast gradient method combined with penalty framework we also derive a total computational complexity of $\mathcal{O}\left(\frac{1}{ε^{3/2}}\right)$ projections onto a simple primal set to attain an $ε-$optimal solution for the original problem.

math.OC↗

arXiv2015

Parallel coordinate descent methods for composite minimization: convergence analysis and error bounds

In this paper we propose a distributed version of a randomized block-coordinate descent method for minimizing the sum of a partially separable smooth convex function and a fully separable non-smooth convex function. Under the assumption of block Lipschitz continuity of the gradient of the smooth function, this method is shown to have a sublinear convergence rate. Linear convergence rate of the method is obtained for the newly introduced class of generalized error bound functions. We prove that the new class of generalized error bound functions encompasses both global/local error bound functions and smooth strongly convex functions. We also show that the theoretical estimates on the convergence rate depend on the number of blocks chosen randomly and a natural measure of separability of the objective function.

math.OC↗

arXiv2015

Complexity certifications of first order inexact Lagrangian methods for general convex programming

In this chapter we derive computational complexity certifications of first order inexact dual methods for solving general smooth constrained convex problems which can arise in real-time applications, such as model predictive control. When it is difficult to project on the primal constraint set described by a collection of general convex functions, we use the Lagrangian relaxation to handle the complicated constraints and then, we apply dual (fast) gradient algorithms based on inexact dual gradient information for solving the corresponding dual problem. The iteration complexity analysis is based on two types of approximate primal solutions: the primal last iterate and an average of primal iterates. We provide sublinear computational complexity estimates on the primal suboptimality and constraint (feasibility) violation of the generated approximate primal solutions. In the final part of the chapter, we present an open-source quadratic optimization solver, referred to as DuQuad, for convex quadratic programs and for evaluation of its behaviour. The solver contains the C-language implementations of the analyzed algorithms.

math.OC↗

arXiv2015

Adaptive inexact fast augmented Lagrangian methods for constrained convex optimization

In this paper we analyze several inexact fast augmented Lagrangian methods for solving linearly constrained convex optimization problems. Mainly, our methods rely on the combination of excessive-gap-like smoothing technique developed in [15] and the newly introduced inexact oracle framework from [4]. We analyze several algorithmic instances with constant and adaptive smoothing parameters and derive total computational complexity results in terms of projections onto a simple primal set. For the basic inexact fast augmented Lagrangian algorithm we obtain the overall computational complexity of order $\mathcal{O}\left(\frac{1}{ε^{5/4}}\right)$, while for the adaptive variant we get $\mathcal{O}\left(\frac{1}ε\right)$, projections onto a primal set in order to obtain an $ε-$optimal solution for our original problem.

math.OC↗

arXiv2015

DuQuad: an inexact (augmented) dual first order algorithm for quadratic programming

In this paper we present the solver DuQuad specialized for solving general convex quadratic problems arising in many engineering applications. When it is difficult to project on the primal feasible set, we use the (augmented) Lagrangian relaxation to handle the complicated constraints and then, we apply dual first order algorithms based on inexact dual gradient information for solving the corresponding dual problem. The iteration complexity analysis is based on two types of approximate primal solutions: the primal last iterate and an average of primal iterates. We provide computational complexity estimates on the primal suboptimality and feasibility violation of the generated approximate primal solutions. Then, these algorithms are implemented in the programming language C in DuQuad, and optimized for low iteration complexity and low memory footprint. DuQuad has a dynamic Matlab interface which make the process of testing, comparing, and analyzing the algorithms simple. The algorithms are implemented using only basic arithmetic and logical operations and are suitable to run on low cost hardware. It is shown that if an approximate solution is sufficient for a given application, there exists problems where some of the implemented algorithms obtain the solution faster than state-of-the-art commercial solvers.

math.OC↗

arXiv2015

Iteration complexity analysis of dual first order methods for conic convex programming

In this paper we provide a detailed analysis of the iteration complexity of dual first order methods for solving conic convex problems. When it is difficult to project on the primal feasible set described by convex constraints, we use the Lagrangian relaxation to handle the complicated constraints and then, we apply dual first order algorithms for solving the corresponding dual problem. We give convergence analysis for dual first order algorithms (dual gradient and fast gradient algorithms): we provide sublinear or linear estimates on the primal suboptimality and feasibility violation of the generated approximate primal solutions. Our analysis relies on the Lipschitz property of the gradient of the dual function or an error bound property of the dual. Furthermore, the iteration complexity analysis is based on two types of approximate primal solutions: the last primal iterate or an average primal sequence.

math.OC↗

arXiv2014

Efficient parallel coordinate descent algorithm for convex optimization problems with separable constraints: application to distributed MPC

In this paper we propose a parallel coordinate descent algorithm for solving smooth convex optimization problems with separable constraints that may arise e.g. in distributed model predictive control (MPC) for linear network systems. Our algorithm is based on block coordinate descent updates in parallel and has a very simple iteration. We prove (sub)linear rate of convergence for the new algorithm under standard assumptions for smooth convex optimization. Further, our algorithm uses local information and thus is suitable for distributed implementations. Moreover, it has low iteration complexity, which makes it appropriate for embedded control. An MPC scheme based on this new parallel algorithm is derived, for which every subsystem in the network can compute feasible and stabilizing control inputs using distributed and cheap computations. For ensuring stability of the MPC scheme, we use a terminal cost formulation derived from a distributed synthesis. Preliminary numerical tests show better performance for our optimization algorithm than other existing methods.

math.OC↗

arXiv2014

On linear convergence of a distributed dual gradient algorithm for linearly constrained separable convex problems

In this paper we propose a distributed dual gradient algorithm for minimizing linearly constrained separable convex problems and analyze its rate of convergence. In particular, we prove that under the assumption of strong convexity and Lipshitz continuity of the gradient of the primal objective function we have a global error bound type property for the dual problem. Using this error bound property we devise a fully distributed dual gradient scheme, i.e. a gradient scheme based on a weighted step size, for which we derive global linear rate of convergence for both dual and primal suboptimality and for primal feasibility violation. Many real applications, e.g. distributed model predictive control, network utility maximization or optimal power flow, can be posed as linearly constrained separable convex problems for which dual gradient type methods from literature have sublinear convergence rate. In the present paper we prove for the first time that in fact we can achieve linear convergence rate for such algorithms when they are used for solving these applications. Numerical simulations are also provided to confirm our theory.

math.OC↗

arXiv2014

Iteration complexity analysis of random coordinate descent methods for $\ell_0$ regularized convex problems

In this paper we analyze a family of general random block coordinate descent methods for the minimization of $\ell_0$ regularized optimization problems, i.e. the objective function is composed of a smooth convex function and the $\ell_0$ regularization. Our family of methods covers particular cases such as random block coordinate gradient descent and random proximal coordinate descent methods. We analyze necessary optimality conditions for this nonconvex $\ell_0$ regularized problem and devise a separation of the set of local minima into restricted classes based on approximation versions of the objective function. We provide a unified analysis of the almost sure convergence for this family of block coordinate descent algorithms and prove that, for each approximation version, the limit points are local minima from the corresponding restricted class of local minimizers. Under the strong convexity assumption, we prove linear convergence in probability for our family of methods.

math.OC↗

arXiv2014

Distributed dual gradient methods and error bound conditions

In this paper we propose distributed dual gradient algorithms for linearly constrained separable convex problems and analyze their rate of convergence under different assumptions. Under the strong convexity assumption on the primal objective function we propose two distributed dual fast gradient schemes for which we prove sublinear rate of convergence for dual suboptimality but also primal suboptimality and feasibility violation for an average primal sequence or for the last generated primal iterate. Under the additional assumption of Lipshitz continuity of the gradient of the primal objective function we prove a global error bound type property for the dual problem and then we analyze a dual gradient scheme for which we derive global linear rate of convergence for both dual and primal suboptimality and primal feasibility violation. We also provide numerical simulations on optimal power flow problems.

math.OC↗

arXiv2013

Distributed model predictive control of leader-follower systems using an interior point method with efficient computations

Standard model predictive control strategies imply the online computation of control inputs at each sampling instance, which traditionally limits this type of control scheme to systems with slow dynamics. This paper focuses on distributed model predictive control for large-scale systems comprised of interacting linear subsystems, where the online computations required for the control input can be distributed amongst them. A model predictive controller based on a distributed interior point method is derived, for which every subsystem in the network can compute stabilizing control inputs using distributed computations. We introduce local terminal sets and cost functions, which together satisfy distributed invariance conditions for the whole system, that guarantees stability of the closed-loop interconnected system. We show that the synthesis of both terminal sets and terminal cost functions can be done in a distributed framework.

math.OC↗