Saved in:
Bibliographic Details
Main Authors: Zhang, Shijun, Zhao, Hongkai, Zhong, Yimin, Zhou, Haomin
Format: Preprint
Published: 2025
Subjects:
Online Access:https://arxiv.org/abs/2502.18959
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866912445499441152
author Zhang, Shijun
Zhao, Hongkai
Zhong, Yimin
Zhou, Haomin
author_facet Zhang, Shijun
Zhao, Hongkai
Zhong, Yimin
Zhou, Haomin
contents The architecture of a neural network and the selection of its activation function are both fundamental to its performance. Equally vital is ensuring these two elements are well-matched, as their alignment is key to achieving effective representation and learning. In this paper, we introduce the Fourier Multi-Component and Multi-Layer Neural Network (FMMNN), a novel model that creates a strong synergy between them. We demonstrate that FMMNNs are highly effective and flexible in modeling high-frequency components. Our theoretical results demonstrate that FMMNNs have exponential expressive power for function approximation. We also analyze the optimization landscape of FMMNNs and find it to be much more favorable than that of standard fully connected neural networks, especially when dealing with high-frequency features. In addition, we propose a scaled random initialization method for the first layer's weights in FMMNNs, which significantly speeds up training and enhances overall performance. Extensive numerical experiments support our theoretical insights, showing that FMMNNs consistently outperform traditional approaches in accuracy and efficiency across various tasks.
format Preprint
id arxiv_https___arxiv_org_abs_2502_18959
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Fourier Multi-Component and Multi-Layer Neural Networks: Unlocking High-Frequency Potential
Zhang, Shijun
Zhao, Hongkai
Zhong, Yimin
Zhou, Haomin
Machine Learning
The architecture of a neural network and the selection of its activation function are both fundamental to its performance. Equally vital is ensuring these two elements are well-matched, as their alignment is key to achieving effective representation and learning. In this paper, we introduce the Fourier Multi-Component and Multi-Layer Neural Network (FMMNN), a novel model that creates a strong synergy between them. We demonstrate that FMMNNs are highly effective and flexible in modeling high-frequency components. Our theoretical results demonstrate that FMMNNs have exponential expressive power for function approximation. We also analyze the optimization landscape of FMMNNs and find it to be much more favorable than that of standard fully connected neural networks, especially when dealing with high-frequency features. In addition, we propose a scaled random initialization method for the first layer's weights in FMMNNs, which significantly speeds up training and enhances overall performance. Extensive numerical experiments support our theoretical insights, showing that FMMNNs consistently outperform traditional approaches in accuracy and efficiency across various tasks.
title Fourier Multi-Component and Multi-Layer Neural Networks: Unlocking High-Frequency Potential
topic Machine Learning
url https://arxiv.org/abs/2502.18959