Nonlinearity Enhanced Adaptive Activation Functions

Fuente: arXiv
Saved in:
Bibliographic Details
Main Author: Yevick, David
Format: Preprint
Published: 2024
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866915284458143744
author Yevick, David
author_facet Yevick, David
contents A general procedure for introducing parametric, learned, nonlinearity into activation functions is found to enhance the accuracy of representative neural networks without requiring significant additional computational resources. Examples are given based on the standard rectified linear unit (ReLU) as well as several other frequently employed activation functions. The associated accuracy improvement is quantified both in the context of the MNIST digit data set and a convolutional neural network (CNN) benchmark example.
format Preprint
id arxiv_https___arxiv_org_abs_2403_19896
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle Nonlinearity Enhanced Adaptive Activation Functions
Yevick, David
Machine Learning
Computer Vision and Pattern Recognition
Neural and Evolutionary Computing
A general procedure for introducing parametric, learned, nonlinearity into activation functions is found to enhance the accuracy of representative neural networks without requiring significant additional computational resources. Examples are given based on the standard rectified linear unit (ReLU) as well as several other frequently employed activation functions. The associated accuracy improvement is quantified both in the context of the MNIST digit data set and a convolutional neural network (CNN) benchmark example.
title Nonlinearity Enhanced Adaptive Activation Functions
topic Machine Learning
Computer Vision and Pattern Recognition
Neural and Evolutionary Computing
url https://arxiv.org/abs/2403.19896