Predicting processes when embedded events occur: Dynamic time warping

Benjamin J. Nelson, George Runger

Research output: Contribution to journalArticle

2 Citations (Scopus)

Abstract

In a traditional time-series analysis one assumes a model for the system. This model has a specified form that is consistent over time. However, in most continuous processes changing tank levels, system upsets (events), and transient process behaviors impact the accuracy of forecasts based on these global models. System behavior is often more closely related to local (short-term) process conditions and events that are repeated at random over time. The global model can predict average behavior, but it often does not accurately predict the local behavior of such a system. The traditional purpose of dynamic time warping was to detect predetermined patterns in speech recognition (time-series data). The technique depends on the development of templates, which represent known patterns or behavior. In this paper we extend the dynamic time warping technique to forecast process measurements using recent process data to define a (dynamic) template. This new dynamic template, dynamic-time-warping approach is compared to traditional methods of predicting process measurements when behavior typical of a continuous process is present. The key properties of such processes are described. The dynamic time warping approach is shown to be an effective method of prediction under these process assumptions.

Original languageEnglish (US)
Pages (from-to)213-226
Number of pages14
JournalJournal of Quality Technology
Volume35
Issue number2
StatePublished - Apr 2003

Fingerprint

Dynamic Time Warping
Template
Time series analysis
Forecast
Speech recognition
Warping
Time series
Predict
Time Series Analysis
Speech Recognition
Time Series Data
Model

Keywords

  • Forecasting techniques
  • Time-series analysis

ASJC Scopus subject areas

  • Industrial and Manufacturing Engineering
  • Statistics and Probability
  • Management Science and Operations Research

Cite this

Predicting processes when embedded events occur : Dynamic time warping. / Nelson, Benjamin J.; Runger, George.

In: Journal of Quality Technology, Vol. 35, No. 2, 04.2003, p. 213-226.

Research output: Contribution to journalArticle

@article{19be1e8ee76f4a65a823f48a3a917e4f,
title = "Predicting processes when embedded events occur: Dynamic time warping",
abstract = "In a traditional time-series analysis one assumes a model for the system. This model has a specified form that is consistent over time. However, in most continuous processes changing tank levels, system upsets (events), and transient process behaviors impact the accuracy of forecasts based on these global models. System behavior is often more closely related to local (short-term) process conditions and events that are repeated at random over time. The global model can predict average behavior, but it often does not accurately predict the local behavior of such a system. The traditional purpose of dynamic time warping was to detect predetermined patterns in speech recognition (time-series data). The technique depends on the development of templates, which represent known patterns or behavior. In this paper we extend the dynamic time warping technique to forecast process measurements using recent process data to define a (dynamic) template. This new dynamic template, dynamic-time-warping approach is compared to traditional methods of predicting process measurements when behavior typical of a continuous process is present. The key properties of such processes are described. The dynamic time warping approach is shown to be an effective method of prediction under these process assumptions.",
keywords = "Forecasting techniques, Time-series analysis",
author = "Nelson, {Benjamin J.} and George Runger",
year = "2003",
month = "4",
language = "English (US)",
volume = "35",
pages = "213--226",
journal = "Journal of Quality Technology",
issn = "0022-4065",
publisher = "American Society for Quality",
number = "2",

}

TY - JOUR

T1 - Predicting processes when embedded events occur

T2 - Dynamic time warping

AU - Nelson, Benjamin J.

AU - Runger, George

PY - 2003/4

Y1 - 2003/4

N2 - In a traditional time-series analysis one assumes a model for the system. This model has a specified form that is consistent over time. However, in most continuous processes changing tank levels, system upsets (events), and transient process behaviors impact the accuracy of forecasts based on these global models. System behavior is often more closely related to local (short-term) process conditions and events that are repeated at random over time. The global model can predict average behavior, but it often does not accurately predict the local behavior of such a system. The traditional purpose of dynamic time warping was to detect predetermined patterns in speech recognition (time-series data). The technique depends on the development of templates, which represent known patterns or behavior. In this paper we extend the dynamic time warping technique to forecast process measurements using recent process data to define a (dynamic) template. This new dynamic template, dynamic-time-warping approach is compared to traditional methods of predicting process measurements when behavior typical of a continuous process is present. The key properties of such processes are described. The dynamic time warping approach is shown to be an effective method of prediction under these process assumptions.

AB - In a traditional time-series analysis one assumes a model for the system. This model has a specified form that is consistent over time. However, in most continuous processes changing tank levels, system upsets (events), and transient process behaviors impact the accuracy of forecasts based on these global models. System behavior is often more closely related to local (short-term) process conditions and events that are repeated at random over time. The global model can predict average behavior, but it often does not accurately predict the local behavior of such a system. The traditional purpose of dynamic time warping was to detect predetermined patterns in speech recognition (time-series data). The technique depends on the development of templates, which represent known patterns or behavior. In this paper we extend the dynamic time warping technique to forecast process measurements using recent process data to define a (dynamic) template. This new dynamic template, dynamic-time-warping approach is compared to traditional methods of predicting process measurements when behavior typical of a continuous process is present. The key properties of such processes are described. The dynamic time warping approach is shown to be an effective method of prediction under these process assumptions.

KW - Forecasting techniques

KW - Time-series analysis

UR - http://www.scopus.com/inward/record.url?scp=1642438105&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=1642438105&partnerID=8YFLogxK

M3 - Article

AN - SCOPUS:1642438105

VL - 35

SP - 213

EP - 226

JO - Journal of Quality Technology

JF - Journal of Quality Technology

SN - 0022-4065

IS - 2

ER -