Analysis of drifting dynamics with neural network hidden markov models

J. Kohlmorgen, Klaus Muller, K. Pawelzik

Research output: Chapter in Book/Report/Conference proceedingConference contribution

3 Citations (Scopus)

Abstract

We present a method for the analysis of nonstationary time series with multiple operating modes. In particular, it is possible to detect and to model both a switching of the dynamics and a less abrupt, time consuming drift from one mode to another. This is achieved in two steps. First, an unsupervised training method provides prediction experts for the inherent dynamical modes. Then, the trained experts are used in a hidden Markov model that allows to model drifts. An application to physiological wake/sleep data demonstrates that analysis and modeling of real-world time series can be improved when the drift paradigm is taken into account.

Original languageEnglish
Title of host publicationAdvances in Neural Information Processing Systems
PublisherNeural information processing systems foundation
Pages735-741
Number of pages7
ISBN (Print)0262100762, 9780262100762
Publication statusPublished - 1998 Jan 1
Externally publishedYes
Event11th Annual Conference on Neural Information Processing Systems, NIPS 1997 - Denver, CO, United States
Duration: 1997 Dec 11997 Dec 6

Other

Other11th Annual Conference on Neural Information Processing Systems, NIPS 1997
CountryUnited States
CityDenver, CO
Period97/12/197/12/6

Fingerprint

Hidden Markov models
Time series
Neural networks
Sleep

ASJC Scopus subject areas

  • Computer Networks and Communications
  • Information Systems
  • Signal Processing

Cite this

Kohlmorgen, J., Muller, K., & Pawelzik, K. (1998). Analysis of drifting dynamics with neural network hidden markov models. In Advances in Neural Information Processing Systems (pp. 735-741). Neural information processing systems foundation.

Analysis of drifting dynamics with neural network hidden markov models. / Kohlmorgen, J.; Muller, Klaus; Pawelzik, K.

Advances in Neural Information Processing Systems. Neural information processing systems foundation, 1998. p. 735-741.

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Kohlmorgen, J, Muller, K & Pawelzik, K 1998, Analysis of drifting dynamics with neural network hidden markov models. in Advances in Neural Information Processing Systems. Neural information processing systems foundation, pp. 735-741, 11th Annual Conference on Neural Information Processing Systems, NIPS 1997, Denver, CO, United States, 97/12/1.
Kohlmorgen J, Muller K, Pawelzik K. Analysis of drifting dynamics with neural network hidden markov models. In Advances in Neural Information Processing Systems. Neural information processing systems foundation. 1998. p. 735-741
Kohlmorgen, J. ; Muller, Klaus ; Pawelzik, K. / Analysis of drifting dynamics with neural network hidden markov models. Advances in Neural Information Processing Systems. Neural information processing systems foundation, 1998. pp. 735-741
@inproceedings{f28e620423f24d9f90ab2240b9626ebb,
title = "Analysis of drifting dynamics with neural network hidden markov models",
abstract = "We present a method for the analysis of nonstationary time series with multiple operating modes. In particular, it is possible to detect and to model both a switching of the dynamics and a less abrupt, time consuming drift from one mode to another. This is achieved in two steps. First, an unsupervised training method provides prediction experts for the inherent dynamical modes. Then, the trained experts are used in a hidden Markov model that allows to model drifts. An application to physiological wake/sleep data demonstrates that analysis and modeling of real-world time series can be improved when the drift paradigm is taken into account.",
author = "J. Kohlmorgen and Klaus Muller and K. Pawelzik",
year = "1998",
month = "1",
day = "1",
language = "English",
isbn = "0262100762",
pages = "735--741",
booktitle = "Advances in Neural Information Processing Systems",
publisher = "Neural information processing systems foundation",

}

TY - GEN

T1 - Analysis of drifting dynamics with neural network hidden markov models

AU - Kohlmorgen, J.

AU - Muller, Klaus

AU - Pawelzik, K.

PY - 1998/1/1

Y1 - 1998/1/1

N2 - We present a method for the analysis of nonstationary time series with multiple operating modes. In particular, it is possible to detect and to model both a switching of the dynamics and a less abrupt, time consuming drift from one mode to another. This is achieved in two steps. First, an unsupervised training method provides prediction experts for the inherent dynamical modes. Then, the trained experts are used in a hidden Markov model that allows to model drifts. An application to physiological wake/sleep data demonstrates that analysis and modeling of real-world time series can be improved when the drift paradigm is taken into account.

AB - We present a method for the analysis of nonstationary time series with multiple operating modes. In particular, it is possible to detect and to model both a switching of the dynamics and a less abrupt, time consuming drift from one mode to another. This is achieved in two steps. First, an unsupervised training method provides prediction experts for the inherent dynamical modes. Then, the trained experts are used in a hidden Markov model that allows to model drifts. An application to physiological wake/sleep data demonstrates that analysis and modeling of real-world time series can be improved when the drift paradigm is taken into account.

UR - http://www.scopus.com/inward/record.url?scp=0006132299&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=0006132299&partnerID=8YFLogxK

M3 - Conference contribution

AN - SCOPUS:0006132299

SN - 0262100762

SN - 9780262100762

SP - 735

EP - 741

BT - Advances in Neural Information Processing Systems

PB - Neural information processing systems foundation

ER -