The Computational Power of Neural Networks and Representations of Numbers in Non-Integer Bases
Loading...
Date
2017-06-01
Authors
Sima, Jiri
ORCID
Advisor
Referee
Mark
Journal Title
Journal ISSN
Volume Title
Publisher
Institute of Automation and Computer Science, Brno University of Technology
Altmetrics
Abstract
We briefly survey the basic concepts and results concerning the computational power of neural net-orks which basically depends on the information content of eight parameters. In particular, recurrent neural networks with integer, rational, and arbitrary real weights are classi ed within the Chomsky and finer complexity hierarchies. Then we re ne the analysis between integer and rational weights by investigating an intermediate model of integer-weight neural networks with an extra analog rational-weight neuron (1ANN). We show a representation theorem which characterizes the classification problems solvable by 1ANNs, by using so-called cut languages. Our analysis reveals an interesting link to an active research field on non-standard positional numeral systems with non-integer bases. Within this framework, we introduce a new concept of quasi-periodic numbers which is used to classify the computational power of 1ANNs within the Chomsky hierarchy.
Description
Citation
Mendel. 2017 vol. 23, č. 1, s. 103-110. ISSN 1803-3814
https://mendel-journal.org/index.php/mendel/article/view/59
https://mendel-journal.org/index.php/mendel/article/view/59
Document type
Peer-reviewed
Document version
Published version
Date of access to the full text
Language of document
en