Time delay estimation with hidden Markov models

Mehdi Azzouzi, Ian T. Nabney

Research output: Chapter in Book/Report/Conference proceedingChapter

Abstract

Most traditional methods for extracting the relationships between two time series are based on cross-correlation. In a non-linear non-stationary environment, these techniques are not sufficient. We show in this paper how to use hidden Markov models to identify the lag (or delay) between different variables for such data. Adopting an information-theoretic approach, we develop a procedure for training HMMs to maximise the mutual information (MMI) between delayed time series. The method is used to model the oil drilling process. We show that cross-correlation gives no information and that the MMI approach outperforms maximum likelihood.
Original languageEnglish
Title of host publicationNinth International Conference on Artificial Neural Networks, 1999 (ICANN)
Place of PublicationEdinburgh, UK
PublisherIEEE
Pages473-478
Number of pages6
Volume1
ISBN (Print)0852967217
Publication statusPublished - 1999
EventNinth International conference on Artificial Neural Networks -
Duration: 1 Jan 19991 Jan 1999

Conference

ConferenceNinth International conference on Artificial Neural Networks
Period1/01/991/01/99

Fingerprint

Hidden Markov models
Time series
Time delay
Maximum likelihood
Drilling
Oils

Bibliographical note

©1999 IEEE. Personal use of this material is permitted. However, permission to reprint/republish this material for advertising or promotional purposes or for creating new collective works for resale or redistribution to servers or lists, or to reuse any copyrighted component of this work in other works must be obtained from the IEEE.

Keywords

  • time series
  • cross-correlation
  • non-stationary
  • Markov models
  • information-theoretic
  • mutual information

Cite this

Azzouzi, M., & Nabney, I. T. (1999). Time delay estimation with hidden Markov models. In Ninth International Conference on Artificial Neural Networks, 1999 (ICANN) (Vol. 1, pp. 473-478). Edinburgh, UK: IEEE.
Azzouzi, Mehdi ; Nabney, Ian T. / Time delay estimation with hidden Markov models. Ninth International Conference on Artificial Neural Networks, 1999 (ICANN). Vol. 1 Edinburgh, UK : IEEE, 1999. pp. 473-478
@inbook{2cabcadddb2d4286803e2f2c0f249af9,
title = "Time delay estimation with hidden Markov models",
abstract = "Most traditional methods for extracting the relationships between two time series are based on cross-correlation. In a non-linear non-stationary environment, these techniques are not sufficient. We show in this paper how to use hidden Markov models to identify the lag (or delay) between different variables for such data. Adopting an information-theoretic approach, we develop a procedure for training HMMs to maximise the mutual information (MMI) between delayed time series. The method is used to model the oil drilling process. We show that cross-correlation gives no information and that the MMI approach outperforms maximum likelihood.",
keywords = "time series, cross-correlation, non-stationary, Markov models, information-theoretic, mutual information",
author = "Mehdi Azzouzi and Nabney, {Ian T.}",
note = "{\circledC}1999 IEEE. Personal use of this material is permitted. However, permission to reprint/republish this material for advertising or promotional purposes or for creating new collective works for resale or redistribution to servers or lists, or to reuse any copyrighted component of this work in other works must be obtained from the IEEE.",
year = "1999",
language = "English",
isbn = "0852967217",
volume = "1",
pages = "473--478",
booktitle = "Ninth International Conference on Artificial Neural Networks, 1999 (ICANN)",
publisher = "IEEE",
address = "United States",

}

Azzouzi, M & Nabney, IT 1999, Time delay estimation with hidden Markov models. in Ninth International Conference on Artificial Neural Networks, 1999 (ICANN). vol. 1, IEEE, Edinburgh, UK, pp. 473-478, Ninth International conference on Artificial Neural Networks, 1/01/99.

Time delay estimation with hidden Markov models. / Azzouzi, Mehdi; Nabney, Ian T.

Ninth International Conference on Artificial Neural Networks, 1999 (ICANN). Vol. 1 Edinburgh, UK : IEEE, 1999. p. 473-478.

Research output: Chapter in Book/Report/Conference proceedingChapter

TY - CHAP

T1 - Time delay estimation with hidden Markov models

AU - Azzouzi, Mehdi

AU - Nabney, Ian T.

N1 - ©1999 IEEE. Personal use of this material is permitted. However, permission to reprint/republish this material for advertising or promotional purposes or for creating new collective works for resale or redistribution to servers or lists, or to reuse any copyrighted component of this work in other works must be obtained from the IEEE.

PY - 1999

Y1 - 1999

N2 - Most traditional methods for extracting the relationships between two time series are based on cross-correlation. In a non-linear non-stationary environment, these techniques are not sufficient. We show in this paper how to use hidden Markov models to identify the lag (or delay) between different variables for such data. Adopting an information-theoretic approach, we develop a procedure for training HMMs to maximise the mutual information (MMI) between delayed time series. The method is used to model the oil drilling process. We show that cross-correlation gives no information and that the MMI approach outperforms maximum likelihood.

AB - Most traditional methods for extracting the relationships between two time series are based on cross-correlation. In a non-linear non-stationary environment, these techniques are not sufficient. We show in this paper how to use hidden Markov models to identify the lag (or delay) between different variables for such data. Adopting an information-theoretic approach, we develop a procedure for training HMMs to maximise the mutual information (MMI) between delayed time series. The method is used to model the oil drilling process. We show that cross-correlation gives no information and that the MMI approach outperforms maximum likelihood.

KW - time series

KW - cross-correlation

KW - non-stationary

KW - Markov models

KW - information-theoretic

KW - mutual information

UR - http://www.scopus.com/inward/record.url?scp=0033323245&partnerID=8YFLogxK

UR - http://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=819766&isnumber=17760

M3 - Chapter

SN - 0852967217

VL - 1

SP - 473

EP - 478

BT - Ninth International Conference on Artificial Neural Networks, 1999 (ICANN)

PB - IEEE

CY - Edinburgh, UK

ER -

Azzouzi M, Nabney IT. Time delay estimation with hidden Markov models. In Ninth International Conference on Artificial Neural Networks, 1999 (ICANN). Vol. 1. Edinburgh, UK: IEEE. 1999. p. 473-478