Learning in Feedforward Neural Networks Accelerated by Transfer Entropy
Current neural networks architectures are many times harder to train because of the increasing size and complexity of the used datasets. Our objective is to design more efficient training algorithms utilizing causal relationships inferred from neural networks. The transfer entropy (TE) was initially...
Main Authors: | , , |
---|---|
Format: | Article |
Language: | English |
Published: |
MDPI AG
2020-01-01
|
Series: | Entropy |
Subjects: | |
Online Access: | https://www.mdpi.com/1099-4300/22/1/102 |
_version_ | 1811299590630014976 |
---|---|
author | Adrian Moldovan Angel Caţaron Răzvan Andonie |
author_facet | Adrian Moldovan Angel Caţaron Răzvan Andonie |
author_sort | Adrian Moldovan |
collection | DOAJ |
description | Current neural networks architectures are many times harder to train because of the increasing size and complexity of the used datasets. Our objective is to design more efficient training algorithms utilizing causal relationships inferred from neural networks. The transfer entropy (TE) was initially introduced as an information transfer measure used to quantify the statistical coherence between events (time series). Later, it was related to causality, even if they are not the same. There are only few papers reporting applications of causality or TE in neural networks. Our contribution is an information-theoretical method for analyzing information transfer between the nodes of feedforward neural networks. The information transfer is measured by the TE of feedback neural connections. Intuitively, TE measures the relevance of a connection in the network and the feedback amplifies this connection. We introduce a backpropagation type training algorithm that uses TE feedback connections to improve its performance. |
first_indexed | 2024-04-13T06:38:07Z |
format | Article |
id | doaj.art-10c11725120c44eb9334c194b75fbb97 |
institution | Directory Open Access Journal |
issn | 1099-4300 |
language | English |
last_indexed | 2024-04-13T06:38:07Z |
publishDate | 2020-01-01 |
publisher | MDPI AG |
record_format | Article |
series | Entropy |
spelling | doaj.art-10c11725120c44eb9334c194b75fbb972022-12-22T02:57:50ZengMDPI AGEntropy1099-43002020-01-0122110210.3390/e22010102e22010102Learning in Feedforward Neural Networks Accelerated by Transfer EntropyAdrian Moldovan0Angel Caţaron1Răzvan Andonie2Department of Electronics and Computers, Transilvania University, 500024 Braşov, RomaniaDepartment of Electronics and Computers, Transilvania University, 500024 Braşov, RomaniaDepartment of Computer Science, Central Washington University, Ellensburg, WA 98926, USACurrent neural networks architectures are many times harder to train because of the increasing size and complexity of the used datasets. Our objective is to design more efficient training algorithms utilizing causal relationships inferred from neural networks. The transfer entropy (TE) was initially introduced as an information transfer measure used to quantify the statistical coherence between events (time series). Later, it was related to causality, even if they are not the same. There are only few papers reporting applications of causality or TE in neural networks. Our contribution is an information-theoretical method for analyzing information transfer between the nodes of feedforward neural networks. The information transfer is measured by the TE of feedback neural connections. Intuitively, TE measures the relevance of a connection in the network and the feedback amplifies this connection. We introduce a backpropagation type training algorithm that uses TE feedback connections to improve its performance.https://www.mdpi.com/1099-4300/22/1/102transfer entropycausalityneural networkbackpropagationgradient descentdeep learning |
spellingShingle | Adrian Moldovan Angel Caţaron Răzvan Andonie Learning in Feedforward Neural Networks Accelerated by Transfer Entropy Entropy transfer entropy causality neural network backpropagation gradient descent deep learning |
title | Learning in Feedforward Neural Networks Accelerated by Transfer Entropy |
title_full | Learning in Feedforward Neural Networks Accelerated by Transfer Entropy |
title_fullStr | Learning in Feedforward Neural Networks Accelerated by Transfer Entropy |
title_full_unstemmed | Learning in Feedforward Neural Networks Accelerated by Transfer Entropy |
title_short | Learning in Feedforward Neural Networks Accelerated by Transfer Entropy |
title_sort | learning in feedforward neural networks accelerated by transfer entropy |
topic | transfer entropy causality neural network backpropagation gradient descent deep learning |
url | https://www.mdpi.com/1099-4300/22/1/102 |
work_keys_str_mv | AT adrianmoldovan learninginfeedforwardneuralnetworksacceleratedbytransferentropy AT angelcataron learninginfeedforwardneuralnetworksacceleratedbytransferentropy AT razvanandonie learninginfeedforwardneuralnetworksacceleratedbytransferentropy |