LGOct 3, 2023

Benign Overfitting in Two-Layer ReLU Convolutional Neural Networks for XOR Data

arXiv:2310.01975v111 citationsh-index: 18
Originality Incremental advance
AI Analysis

This work addresses the theoretical understanding of why overfitting neural networks can still perform well, specifically for non-linear XOR problems with noise, which is incremental as it extends prior analyses beyond linear classifiers.

The paper tackles the problem of benign overfitting in over-parameterized neural networks by analyzing XOR-type classification tasks with label-flipping noise, showing that a ReLU CNN trained by gradient descent can achieve near Bayes-optimal accuracy under specific conditions on sample complexity and signal-to-noise ratio, with a matching lower bound when conditions are not met.

Modern deep learning models are usually highly over-parameterized so that they can overfit the training data. Surprisingly, such overfitting neural networks can usually still achieve high prediction accuracy. To study this "benign overfitting" phenomenon, a line of recent works has theoretically studied the learning of linear models and two-layer neural networks. However, most of these analyses are still limited to the very simple learning problems where the Bayes-optimal classifier is linear. In this work, we investigate a class of XOR-type classification tasks with label-flipping noises. We show that, under a certain condition on the sample complexity and signal-to-noise ratio, an over-parameterized ReLU CNN trained by gradient descent can achieve near Bayes-optimal accuracy. Moreover, we also establish a matching lower bound result showing that when the previous condition is not satisfied, the prediction accuracy of the obtained CNN is an absolute constant away from the Bayes-optimal rate. Our result demonstrates that CNNs have a remarkable capacity to efficiently learn XOR problems, even in the presence of highly correlated features.

Foundations

The foundational work for this paper's niche, ranked by how specifically the neighbourhood builds on it — not by global fame.

Your Notes