arXiv subjects

Courtney Paquette

Publications and source records attributed to Courtney Paquette.

25 records · Page 2Linked to original sources

arXiv2020

A termination criterion for stochastic gradient descent for binary classification

We propose a new, simple, and computationally inexpensive termination test for constant step-size stochastic gradient descent (SGD) applied to binary classification on the logistic and hinge loss with homogeneous linear predictors. Our theoretical results support the effectiveness of our stopping criterion when the data is Gaussian distributed. This presence of noise allows for the possibility of non-separable data. We show that our test terminates in a finite number of iterations and when the noise in the data is not too large, the expected classifier at termination nearly minimizes the probability of misclassification. Finally, numerical experiments indicate for both real and synthetic data sets that our termination test exhibits a good degree of predictability on accuracy and running time.

math.OC↗

arXiv2019

Potential-based analyses of first-order methods for constrained and composite optimization

We propose potential-based analyses for first-order algorithms applied to constrained and composite minimization problems. We first propose ``idealized'' frameworks for algorithms in the strongly and non-strongly convex cases and argue based on a potential that methods following the framework achieve the best possible rate. Then we show that the geometric descent (GD) algorithm by Bubeck et al.\ as extended to the constrained and composite setting by Chen et al.\ achieves this rate using the potential-based analysis for the strongly convex case. Next, we extend the GD algorithm to the case of non-strongly convex problems. We show using a related potential-based argument that our extension achieves the best possible rate in this case as well. The new GD algorithm achieves the best possible rate in the nonconvex case also. We also analyze accelerated gradient using the new potentials. We then turn to the special case of a quadratic function with a single ball constraint, the famous trust-region subproblem. For this case, the first-order trust-region Lanczos method by Gould et al.\ finds the optimal point in an increasing sequence of Krylov spaces. Our results for the general case immediately imply convergence rates for their method in both the strongly convex and non-strongly convex cases. We also establish the same convergence rates for their method using arguments based on Chebyshev polynomial approximation. To the best of our knowledge, no convergence rate has previously been established for the trust-region Lanczos method.

math.OC↗

arXiv2018

Catalyst Acceleration for Gradient-Based Non-Convex Optimization

We introduce a generic scheme to solve nonconvex optimization problems using gradient-based algorithms originally designed for minimizing convex functions. Even though these methods may originally require convexity to operate, the proposed approach allows one to use them on weakly convex objectives, which covers a large class of non-convex functions typically appearing in machine learning and signal processing. In general, the scheme is guaranteed to produce a stationary point with a worst-case efficiency typical of first-order methods, and when the objective turns out to be convex, it automatically accelerates in the sense of Nesterov and achieves near-optimal convergence rate in function values. These properties are achieved without assuming any knowledge about the convexity of the objective, by automatically adapting to the unknown weak convexity constant. We conclude the paper by showing promising experimental results obtained by applying our approach to incremental algorithms such as SVRG and SAGA for sparse matrix factorization and for learning neural networks.

stat.ML↗

arXiv2018

A Stochastic Line Search Method with Convergence Rate Analysis

For deterministic optimization, line-search methods augment algorithms by providing stability and improved efficiency. We adapt a classical backtracking Armijo line-search to the stochastic optimization setting. While traditional line-search relies on exact computations of the gradient and values of the objective function, our method assumes that these values are available up to some dynamically adjusted accuracy which holds with some sufficiently large, but fixed, probability. We show the expected number of iterations to reach a near stationary point matches the worst-case efficiency of typical first-order methods, while for convex and strongly convex objective, it achieves rates of deterministic gradient descent in function values.

math.OC↗

arXiv2018

Subgradient methods for sharp weakly convex functions

Subgradient methods converge linearly on a convex function that grows sharply away from its solution set. In this work, we show that the same is true for sharp functions that are only weakly convex, provided that the subgradient methods are initialized within a fixed tube around the solution set. A variety of statistical and signal processing tasks come equipped with good initialization, and provably lead to formulations that are both weakly convex and sharp. Therefore, in such settings, subgradient methods can serve as inexpensive local search procedures. We illustrate the proposed techniques on phase retrieval and covariance estimation problems.

math.OC↗

arXiv2018

The nonsmooth landscape of phase retrieval

We consider a popular nonsmooth formulation of the real phase retrieval problem. We show that under standard statistical assumptions, a simple subgradient method converges linearly when initialized within a constant relative distance of an optimal solution. Seeking to understand the distribution of the stationary points of the problem, we complete the paper by proving that as the number of Gaussian measurements increases, the stationary points converge to a codimension two set, at a controlled rate. Experiments on image recovery problems illustrate the developed algorithm and theory.

math.OC↗

arXiv2017

Efficiency of minimizing compositions of convex functions and smooth maps

We consider global efficiency of algorithms for minimizing a sum of a convex function and a composition of a Lipschitz convex function with a smooth map. The basic algorithm we rely on is the prox-linear method, which in each iteration solves a regularized subproblem formed by linearizing the smooth map. When the subproblems are solved exactly, the method has efficiency $\mathcal{O}(\varepsilon^{-2})$, akin to gradient descent for smooth minimization. We show that when the subproblems can only be solved by first-order methods, a simple combination of smoothing, the prox-linear method, and a fast-gradient scheme yields an algorithm with complexity $\widetilde{\mathcal{O}}(\varepsilon^{-3})$. The technique readily extends to minimizing an average of $m$ composite functions, with complexity $\widetilde{\mathcal{O}}(m/\varepsilon^{2}+\sqrt{m}/\varepsilon^{3})$ in expectation. We round off the paper with an inertial prox-linear method that automatically accelerates in presence of convexity.

math.OC↗