Source author record

Anatoli Iouditski

Explore this author’s publications, collaborators and research topics. This catalog record has not been claimed by its author.

ResearcherUnclaimed source record

math.OC math.ST Methodology Statistics Theory Systems and Control

Catalog footprint

Research overview

5works

5topics

4close collaborators

Actions

Is this your work?

Open graph Browse papers

preprint2014arXiv

Primal-dual subgradient methods for minimizing uniformly convex functions

We discuss non-Euclidean deterministic and stochastic algorithms for optimization problems with strongly and uniformly convex objectives. We provide accuracy bounds for the performance of these algorithms and design methods which are adaptive with respect to the parameters of strong or uniform convexity of the objective: in the case when the total number of iterations $N$ is fixed, their accuracy coincides, up to a logarithmic in $N$ factor with the accuracy of optimal algorithms.

preprint2011arXiv

Accuracy guarantees for L1-recovery

We discuss two new methods of recovery of sparse signals from noisy observation based on $\ell_1$ - minimization. They are closely related to the well-known techniques such as Lasso and Dantzig Selector. However, these estimators come with efficiently verifiable guaranties of performance. By optimizing these bounds with respect to the method parameters we are able to construct the estimators which possess better statistical properties than the commonly used ones. We also show how these techniques allow to provide efficiently computable accuracy bounds for Lasso and Dantzig Selector. We link our performance estimations to the well known results of Compressive Sensing and justify our proposed approach with an oracle inequality which links the properties of the recovery algorithms and the best estimation performance when the signal support is known. We demonstrate how the estimates can be computed using the Non-Euclidean Basis Pursuit algorithm.

preprint2011arXiv

Linear programming problems for frontier estimation

We propose new estimates for the frontier of a set of points. They are defined as kernel estimates covering all the points and whose associated support is of smallest surface. The estimates are written as linear combinatio- ns of kernel functions applied to the points of the sample. The coefficients of the linear combination are then computed by solving a linear programming problem. In the general case, the solution of the optimizat- ion problem is sparse, that is, only a few coefficients are non zero. The corresponding points play the role of support vectors in the statistical learning theory. The L_1 error between the estimated and the true frontiers is shown to be almost surely converging to zero, and the rate of convergence is provided. The behaviour of the estimates on finite sample situations is illustrated on some simulations.

preprint2011arXiv

Linear programming problems for l_1- optimal frontier estimation

We propose new optimal estimators for the Lipschitz frontier of a set of points. They are defined as kernel estimators being sufficiently regular, covering all the points and whose associated support is of smallest surface. The estimators are written as linear combinations of kernel functions applied to the points of the sample. The coefficients of the linear combination are then computed by solving related linear programming problem. The L_1 error between the estimated and the true frontier function with a known Lipschitz constant is shown to be almost surely converging to zero, and the rate of convergence is proved to be optimal.

preprint2010arXiv

Verifiable conditions of $\ell_1$ -recovery of sparse signals with sign restrictions

We propose necessary and sufficient conditions for a sensing matrix to be "s-semigood" -- to allow for exact $\ell_1$ -recovery of sparse signals with at most $s$ nonzero entries under sign restrictions on part of the entries. We express the error bounds for imperfect $\ell_1$ -recovery in terms of the characteristics underlying these conditions. Furthermore, we demonstrate that these characteristics, although difficult to evaluate, lead to verifiable sufficient conditions for exact sparse $\ell_1$ -recovery and to efficiently computable upper bounds on those $s$ for which a given sensing matrix is $s$ -semigood. We concentrate on the properties of proposed verifiable sufficient conditions of $s$ -semigoodness and describe their limits of performance.

Inspect adjacent papers, topics, institutions and collaborators without losing the researcher page.

Building this map preview

BZPEER is loading the nearby papers, people, topics and institutions for this page.

Anatoli Iouditski

Research overview

Is this your work?

5 publications

Primal-dual subgradient methods for minimizing uniformly convex functions

Accuracy guarantees for L1-recovery

Linear programming problems for frontier estimation

Linear programming problems for l_1- optimal frontier estimation

Verifiable conditions of $\ell_1$ -recovery of sparse signals with sign restrictions

See the researcher in context

Building this map preview

Anatoli Iouditski

Research overview

Is this your work?

5 publications

Primal-dual subgradient methods for minimizing uniformly convex functions

Accuracy guarantees for L1-recovery

Linear programming problems for frontier estimation

Linear programming problems for l_1- optimal frontier estimation

Verifiable conditions of ℓ1\ell_1ℓ1​-recovery of sparse signals with sign restrictions

See the researcher in context

Building this map preview

Verifiable conditions of $\ell_1$ -recovery of sparse signals with sign restrictions