Estimation of Deep Neural Networks Capabilities Using Polynomial Approach

Currently very popular trend in artificial intelligence is the use of deep neural networks. The power of such networks are very large, but the main difficulty is learning these networks. The article presents a analysis of deep neural network nonlinearity with polynomial approximation of neuron activ...

Full description

Saved in:
Bibliographic Details
Published inArtificial Intelligence and Soft Computing pp. 136 - 147
Main Authors Rozycki, Pawel, Kolbusz, Janusz, Korostenskyi, Roman, Wilamowski, Bogdan M.
Format Book Chapter
LanguageEnglish
Published Cham Springer International Publishing
SeriesLecture Notes in Computer Science
Subjects
Activation function
Deep neural networks
Nonlinearity
Online AccessGet full text

Cover

More Information
Summary:Currently very popular trend in artificial intelligence is the use of deep neural networks. The power of such networks are very large, but the main difficulty is learning these networks. The article presents a analysis of deep neural network nonlinearity with polynomial approximation of neuron activation functions. It is shown that nonlinearity grows exponentially with the depth of the neural network. The effectiveness of the approach is demonstrated by several experiments.
Bibliography:This work was supported by the National Science Centre, Cracow, Poland under Grant No. 2013/11/B/ST6/01337.
ISBN:9783319393773
3319393774
ISSN:0302-9743
1611-3349
DOI:10.1007/978-3-319-39378-0_13