Saved in:
Bibliographic Details
Main Author: Chen, Shengjian
Format: Preprint
Published: 2025
Subjects:
Online Access:https://arxiv.org/abs/2507.03885
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866918157750370304
author Chen, Shengjian
author_facet Chen, Shengjian
contents We point out that neural networks are not black boxes, and their generalization stems from the ability to dynamically map a dataset to the extrema of the model function. We further prove that the number of extrema in a neural network is positively correlated with the number of its parameters. We then propose a new algorithm that is significantly different from back-propagation algorithm, which mainly obtains the values of parameters by solving a system of linear equations. Some difficult situations, such as gradient vanishing and overfitting, can be simply explained and dealt with in this framework.
format Preprint
id arxiv_https___arxiv_org_abs_2507_03885
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Unraveling the Black Box of Neural Networks: A Dynamic Extremum Mapper
Chen, Shengjian
Machine Learning
We point out that neural networks are not black boxes, and their generalization stems from the ability to dynamically map a dataset to the extrema of the model function. We further prove that the number of extrema in a neural network is positively correlated with the number of its parameters. We then propose a new algorithm that is significantly different from back-propagation algorithm, which mainly obtains the values of parameters by solving a system of linear equations. Some difficult situations, such as gradient vanishing and overfitting, can be simply explained and dealt with in this framework.
title Unraveling the Black Box of Neural Networks: A Dynamic Extremum Mapper
topic Machine Learning
url https://arxiv.org/abs/2507.03885