Saved in:
Bibliographic Details
Main Authors: Berlyand, Leonid, Sandier, Etienne, Shmalo, Yitzchak, Zhang, Lei
Format: Preprint
Published: 2023
Subjects:
Online Access:https://arxiv.org/abs/2310.03165
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866914942231248896
author Berlyand, Leonid
Sandier, Etienne
Shmalo, Yitzchak
Zhang, Lei
author_facet Berlyand, Leonid
Sandier, Etienne
Shmalo, Yitzchak
Zhang, Lei
contents We explore the applications of random matrix theory (RMT) in the training of deep neural networks (DNNs), focusing on layer pruning that is reducing the number of DNN parameters (weights). Our numerical results show that this pruning leads to a drastic reduction of parameters while not reducing the accuracy of DNNs and CNNs. Moreover, pruning the fully connected DNNs actually increases the accuracy and decreases the variance for random initializations. Our numerics indicate that this enhancement in accuracy is due to the simplification of the loss landscape. We next provide rigorous mathematical underpinning of these numerical results by proving the RMT-based Pruning Theorem. Our results offer valuable insights into the practical application of RMT for the creation of more efficient and accurate deep-learning models.
format Preprint
id arxiv_https___arxiv_org_abs_2310_03165
institution arXiv
publishDate 2023
record_format arxiv
spellingShingle Enhancing Accuracy in Deep Learning Using Random Matrix Theory
Berlyand, Leonid
Sandier, Etienne
Shmalo, Yitzchak
Zhang, Lei
Machine Learning
Optimization and Control
We explore the applications of random matrix theory (RMT) in the training of deep neural networks (DNNs), focusing on layer pruning that is reducing the number of DNN parameters (weights). Our numerical results show that this pruning leads to a drastic reduction of parameters while not reducing the accuracy of DNNs and CNNs. Moreover, pruning the fully connected DNNs actually increases the accuracy and decreases the variance for random initializations. Our numerics indicate that this enhancement in accuracy is due to the simplification of the loss landscape. We next provide rigorous mathematical underpinning of these numerical results by proving the RMT-based Pruning Theorem. Our results offer valuable insights into the practical application of RMT for the creation of more efficient and accurate deep-learning models.
title Enhancing Accuracy in Deep Learning Using Random Matrix Theory
topic Machine Learning
Optimization and Control
url https://arxiv.org/abs/2310.03165