English

Outlier Detection Using Generative Models with Theoretical Performance Guarantees

Machine Learning 2023-10-17 v1 Machine Learning Signal Processing

Abstract

This paper considers the problem of recovering signals modeled by generative models from linear measurements contaminated with sparse outliers. We propose an outlier detection approach for reconstructing the ground-truth signals modeled by generative models under sparse outliers. We establish theoretical recovery guarantees for reconstruction of signals using generative models in the presence of outliers, giving lower bounds on the number of correctable outliers. Our results are applicable to both linear generator neural networks and the nonlinear generator neural networks with an arbitrary number of layers. We propose an iterative alternating direction method of multipliers (ADMM) algorithm for solving the outlier detection problem via 1\ell_1 norm minimization, and a gradient descent algorithm for solving the outlier detection problem via squared 1\ell_1 norm minimization. We conduct extensive experiments using variational auto-encoder and deep convolutional generative adversarial networks, and the experimental results show that the signals can be successfully reconstructed under outliers using our approach. Our approach outperforms the traditional Lasso and 2\ell_2 minimization approach.

Keywords

Cite

@article{arxiv.2310.09999,
  title  = {Outlier Detection Using Generative Models with Theoretical Performance Guarantees},
  author = {Jirong Yi and Jingchao Gao and Tianming Wang and Xiaodong Wu and Weiyu Xu},
  journal= {arXiv preprint arXiv:2310.09999},
  year   = {2023}
}

Comments

arXiv admin note: substantial text overlap with arXiv:1810.11335

R2 v1 2026-06-28T12:51:21.835Z