Analysis of switching dynamics with competing neural networks

Klaus Muller, Jens Kohlmorgen, Klaus Pawelzik

Research output: Contribution to journalArticle

22 Citations (Scopus)

Abstract

We present a framework for the unsupervised segmentation of time series. It applies to non-stationary signals originating from different dynamical systems which alternate in time, a phenomenon which appears in many natural systems. In our approach, predictors compete for data points of a given time series. We combine competition and evolutionary inertia to a learning rule. Under this learning rule the system evolves such that the predictors, which finally survive, unambiguously identify the underlying processes. The segmentation achieved by this method is very precise and transients are included, a fact, which makes our approach promising for future applications.

Original languageEnglish
Pages (from-to)1306-1315
Number of pages10
JournalIEICE Transactions on Fundamentals of Electronics, Communications and Computer Sciences
VolumeE78-A
Issue number10
Publication statusPublished - 1995 Oct 1
Externally publishedYes

Fingerprint

Rule Learning
Time series
Predictors
Segmentation
Neural Networks
Neural networks
Non-stationary Signal
Alternate
Inertia
Dynamical systems
Dynamical system
Framework

ASJC Scopus subject areas

  • Hardware and Architecture
  • Information Systems
  • Electrical and Electronic Engineering

Cite this

Analysis of switching dynamics with competing neural networks. / Muller, Klaus; Kohlmorgen, Jens; Pawelzik, Klaus.

In: IEICE Transactions on Fundamentals of Electronics, Communications and Computer Sciences, Vol. E78-A, No. 10, 01.10.1995, p. 1306-1315.

Research output: Contribution to journalArticle

Muller, Klaus ; Kohlmorgen, Jens ; Pawelzik, Klaus. / Analysis of switching dynamics with competing neural networks. In: IEICE Transactions on Fundamentals of Electronics, Communications and Computer Sciences. 1995 ; Vol. E78-A, No. 10. pp. 1306-1315.
@article{655658c0fbbb43ec89a4521edbca3454,
title = "Analysis of switching dynamics with competing neural networks",
abstract = "We present a framework for the unsupervised segmentation of time series. It applies to non-stationary signals originating from different dynamical systems which alternate in time, a phenomenon which appears in many natural systems. In our approach, predictors compete for data points of a given time series. We combine competition and evolutionary inertia to a learning rule. Under this learning rule the system evolves such that the predictors, which finally survive, unambiguously identify the underlying processes. The segmentation achieved by this method is very precise and transients are included, a fact, which makes our approach promising for future applications.",
author = "Klaus Muller and Jens Kohlmorgen and Klaus Pawelzik",
year = "1995",
month = "10",
day = "1",
language = "English",
volume = "E78-A",
pages = "1306--1315",
journal = "IEICE Transactions on Fundamentals of Electronics, Communications and Computer Sciences",
issn = "0916-8508",
publisher = "Maruzen Co., Ltd/Maruzen Kabushikikaisha",
number = "10",

}

TY - JOUR

T1 - Analysis of switching dynamics with competing neural networks

AU - Muller, Klaus

AU - Kohlmorgen, Jens

AU - Pawelzik, Klaus

PY - 1995/10/1

Y1 - 1995/10/1

N2 - We present a framework for the unsupervised segmentation of time series. It applies to non-stationary signals originating from different dynamical systems which alternate in time, a phenomenon which appears in many natural systems. In our approach, predictors compete for data points of a given time series. We combine competition and evolutionary inertia to a learning rule. Under this learning rule the system evolves such that the predictors, which finally survive, unambiguously identify the underlying processes. The segmentation achieved by this method is very precise and transients are included, a fact, which makes our approach promising for future applications.

AB - We present a framework for the unsupervised segmentation of time series. It applies to non-stationary signals originating from different dynamical systems which alternate in time, a phenomenon which appears in many natural systems. In our approach, predictors compete for data points of a given time series. We combine competition and evolutionary inertia to a learning rule. Under this learning rule the system evolves such that the predictors, which finally survive, unambiguously identify the underlying processes. The segmentation achieved by this method is very precise and transients are included, a fact, which makes our approach promising for future applications.

UR - http://www.scopus.com/inward/record.url?scp=0029389427&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=0029389427&partnerID=8YFLogxK

M3 - Article

VL - E78-A

SP - 1306

EP - 1315

JO - IEICE Transactions on Fundamentals of Electronics, Communications and Computer Sciences

JF - IEICE Transactions on Fundamentals of Electronics, Communications and Computer Sciences

SN - 0916-8508

IS - 10

ER -