Learning and predicting time series by neural networks

Ansgar Freking, Wolfgang Kinzel, Ido Kanter

Research output: Contribution to journalArticlepeer-review

2 Scopus citations

Abstract

Artificial neural networks which are trained on a time series are supposed to achieve two abilities: first, to predict the series many time steps ahead and second, to learn the rule which has produced the series. It is shown that prediction and learning are not necessarily related to each other. Chaotic sequences can be learned but not predicted while quasiperiodic sequences can be well predicted but not learned.

Original languageEnglish
Pages (from-to)50903
Number of pages1
JournalPhysical Review E
Volume65
Issue number5
StatePublished - 1 May 2002
Externally publishedYes

Fingerprint

Dive into the research topics of 'Learning and predicting time series by neural networks'. Together they form a unique fingerprint.

Cite this