Saved in:
Bibliographic Details
Main Authors: Pfeiffer, Pia, Alfons, Andreas, Filzmoser, Peter
Format: Preprint
Published: 2023
Subjects:
Online Access:https://arxiv.org/abs/2311.17563
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866917907002294272
author Pfeiffer, Pia
Alfons, Andreas
Filzmoser, Peter
author_facet Pfeiffer, Pia
Alfons, Andreas
Filzmoser, Peter
contents Robust statistical estimators offer resilience against outliers but are often computationally challenging, particularly in high-dimensional sparse settings. Modern optimization techniques are utilized for robust sparse association estimators without imposing constraints on the covariance structure. The approach splits the problem into a robust estimation phase, followed by optimization of a decoupled, biconvex problem to derive the sparse canonical vectors. An augmented Lagrangian algorithm, combined with a modified adaptive gradient descent method, induces sparsity through simultaneous updates of both canonical vectors. Results demonstrate improved precision over existing methods, with high-dimensional empirical examples illustrating the effectiveness of this approach. The methodology can also be extended to other robust sparse estimators.
format Preprint
id arxiv_https___arxiv_org_abs_2311_17563
institution arXiv
publishDate 2023
record_format arxiv
spellingShingle Efficient Computation of Sparse and Robust Maximum Association Estimators
Pfeiffer, Pia
Alfons, Andreas
Filzmoser, Peter
Computation
Machine Learning
Robust statistical estimators offer resilience against outliers but are often computationally challenging, particularly in high-dimensional sparse settings. Modern optimization techniques are utilized for robust sparse association estimators without imposing constraints on the covariance structure. The approach splits the problem into a robust estimation phase, followed by optimization of a decoupled, biconvex problem to derive the sparse canonical vectors. An augmented Lagrangian algorithm, combined with a modified adaptive gradient descent method, induces sparsity through simultaneous updates of both canonical vectors. Results demonstrate improved precision over existing methods, with high-dimensional empirical examples illustrating the effectiveness of this approach. The methodology can also be extended to other robust sparse estimators.
title Efficient Computation of Sparse and Robust Maximum Association Estimators
topic Computation
Machine Learning
url https://arxiv.org/abs/2311.17563