Generalization Error Bounds for Iterative Recovery Algorithms Unfolded as Neural Networks

TL;DR

This paper introduces a unified neural network framework for sparse signal reconstruction, deriving generalization bounds that depend linearly on parameters and depth, applicable to various architectures including LISTA and CNNs.

Contribution

It provides a general analysis of neural networks for sparse recovery, encompassing diverse architectures and deriving sample complexity bounds based on Rademacher complexity.

Findings

Generalization bounds depend linearly on network parameters and depth.

Applicable to various architectures like LISTA, dictionary learning, and CNNs.

Provides theoretical guarantees for sparse reconstruction neural networks.

Abstract

Motivated by the learned iterative soft thresholding algorithm (LISTA), we introduce a general class of neural networks suitable for sparse reconstruction from few linear measurements. By allowing a wide range of degrees of weight-sharing between the layers, we enable a unified analysis for very different neural network types, ranging from recurrent ones to networks more similar to standard feedforward neural networks. Based on training samples, via empirical risk minimization we aim at learning the optimal network parameters and thereby the optimal network that reconstructs signals from their low-dimensional linear measurements. We derive generalization bounds by analyzing the Rademacher complexity of hypothesis classes consisting of such deep networks, that also take into account the thresholding parameters. We obtain estimates of the sample complexity that essentially depend only linearly on the number of parameters and on the depth. We apply our main result to obtain specific generalization bounds for several practical examples, including different algorithms for (implicit) dictionary learning, and convolutional neural networks.