Spatiotemporal bio surveillance under non-homogeneous population

Sung Won Han, Wei Jiang, Kwok Leung Tsui

Research output: Contribution to conferencePaper

1 Citation (Scopus)

Abstract

Motivated by the applications in healthcare surveillance, this paper discusses the spatiotemporal surveillance problem of detecting the mean change of Poisson count data in a non-homogeneous population environment. Through Monte Carlo simulations, we investigate several likelihood ratio-based approaches and compare them under various scenarios depending on four factors (1) the population trend, (2) the change time, (3) the change magnitude, and (4) the change coverage. Most literature of spatiotemporal surveillance evaluated the performance based on the average run length if a change occurs at the beginning of surveillance, which is often noted by ARL1. On the other hand, our comparison is based on the average run length after the time when a change occurs later. Our simulation study shows that no method is uniformly better than others in all scenarios. It is found that the difference between generalized likelihood ratios (GLR) approach and weighted likelihood ratios (WLR) approach depends on population trend and change time, not the change coverage or change magnitude.

Original languageEnglish
Pages143-157
Number of pages15
DOIs
Publication statusPublished - 2012 Jan 1
Externally publishedYes
Event2010 10th International Workshop on Intelligent Statistical Quality Control - Seattle, WA, United States
Duration: 2010 Aug 182010 Aug 20

Conference

Conference2010 10th International Workshop on Intelligent Statistical Quality Control
CountryUnited States
CitySeattle, WA
Period10/8/1810/8/20

Fingerprint

Monte Carlo simulation

Keywords

  • Change point detection
  • Clusters
  • Detection delay
  • Generalized likelihood ratios
  • Non-homogeneous poisson
  • Scan statistics
  • Spatiotemporal surveillance
  • Weighted likelihood ratios

ASJC Scopus subject areas

  • Computer Networks and Communications

Cite this

Han, S. W., Jiang, W., & Tsui, K. L. (2012). Spatiotemporal bio surveillance under non-homogeneous population. 143-157. Paper presented at 2010 10th International Workshop on Intelligent Statistical Quality Control, Seattle, WA, United States. https://doi.org/10.1007/978-3-7908-2846-7-11

Spatiotemporal bio surveillance under non-homogeneous population. / Han, Sung Won; Jiang, Wei; Tsui, Kwok Leung.

2012. 143-157 Paper presented at 2010 10th International Workshop on Intelligent Statistical Quality Control, Seattle, WA, United States.

Research output: Contribution to conferencePaper

Han, SW, Jiang, W & Tsui, KL 2012, 'Spatiotemporal bio surveillance under non-homogeneous population' Paper presented at 2010 10th International Workshop on Intelligent Statistical Quality Control, Seattle, WA, United States, 10/8/18 - 10/8/20, pp. 143-157. https://doi.org/10.1007/978-3-7908-2846-7-11
Han SW, Jiang W, Tsui KL. Spatiotemporal bio surveillance under non-homogeneous population. 2012. Paper presented at 2010 10th International Workshop on Intelligent Statistical Quality Control, Seattle, WA, United States. https://doi.org/10.1007/978-3-7908-2846-7-11
Han, Sung Won ; Jiang, Wei ; Tsui, Kwok Leung. / Spatiotemporal bio surveillance under non-homogeneous population. Paper presented at 2010 10th International Workshop on Intelligent Statistical Quality Control, Seattle, WA, United States.15 p.
@conference{5b433bf12d2940fc80972e6a1907b414,
title = "Spatiotemporal bio surveillance under non-homogeneous population",
abstract = "Motivated by the applications in healthcare surveillance, this paper discusses the spatiotemporal surveillance problem of detecting the mean change of Poisson count data in a non-homogeneous population environment. Through Monte Carlo simulations, we investigate several likelihood ratio-based approaches and compare them under various scenarios depending on four factors (1) the population trend, (2) the change time, (3) the change magnitude, and (4) the change coverage. Most literature of spatiotemporal surveillance evaluated the performance based on the average run length if a change occurs at the beginning of surveillance, which is often noted by ARL1. On the other hand, our comparison is based on the average run length after the time when a change occurs later. Our simulation study shows that no method is uniformly better than others in all scenarios. It is found that the difference between generalized likelihood ratios (GLR) approach and weighted likelihood ratios (WLR) approach depends on population trend and change time, not the change coverage or change magnitude.",
keywords = "Change point detection, Clusters, Detection delay, Generalized likelihood ratios, Non-homogeneous poisson, Scan statistics, Spatiotemporal surveillance, Weighted likelihood ratios",
author = "Han, {Sung Won} and Wei Jiang and Tsui, {Kwok Leung}",
year = "2012",
month = "1",
day = "1",
doi = "10.1007/978-3-7908-2846-7-11",
language = "English",
pages = "143--157",
note = "2010 10th International Workshop on Intelligent Statistical Quality Control ; Conference date: 18-08-2010 Through 20-08-2010",

}

TY - CONF

T1 - Spatiotemporal bio surveillance under non-homogeneous population

AU - Han, Sung Won

AU - Jiang, Wei

AU - Tsui, Kwok Leung

PY - 2012/1/1

Y1 - 2012/1/1

N2 - Motivated by the applications in healthcare surveillance, this paper discusses the spatiotemporal surveillance problem of detecting the mean change of Poisson count data in a non-homogeneous population environment. Through Monte Carlo simulations, we investigate several likelihood ratio-based approaches and compare them under various scenarios depending on four factors (1) the population trend, (2) the change time, (3) the change magnitude, and (4) the change coverage. Most literature of spatiotemporal surveillance evaluated the performance based on the average run length if a change occurs at the beginning of surveillance, which is often noted by ARL1. On the other hand, our comparison is based on the average run length after the time when a change occurs later. Our simulation study shows that no method is uniformly better than others in all scenarios. It is found that the difference between generalized likelihood ratios (GLR) approach and weighted likelihood ratios (WLR) approach depends on population trend and change time, not the change coverage or change magnitude.

AB - Motivated by the applications in healthcare surveillance, this paper discusses the spatiotemporal surveillance problem of detecting the mean change of Poisson count data in a non-homogeneous population environment. Through Monte Carlo simulations, we investigate several likelihood ratio-based approaches and compare them under various scenarios depending on four factors (1) the population trend, (2) the change time, (3) the change magnitude, and (4) the change coverage. Most literature of spatiotemporal surveillance evaluated the performance based on the average run length if a change occurs at the beginning of surveillance, which is often noted by ARL1. On the other hand, our comparison is based on the average run length after the time when a change occurs later. Our simulation study shows that no method is uniformly better than others in all scenarios. It is found that the difference between generalized likelihood ratios (GLR) approach and weighted likelihood ratios (WLR) approach depends on population trend and change time, not the change coverage or change magnitude.

KW - Change point detection

KW - Clusters

KW - Detection delay

KW - Generalized likelihood ratios

KW - Non-homogeneous poisson

KW - Scan statistics

KW - Spatiotemporal surveillance

KW - Weighted likelihood ratios

UR - http://www.scopus.com/inward/record.url?scp=84896312002&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84896312002&partnerID=8YFLogxK

U2 - 10.1007/978-3-7908-2846-7-11

DO - 10.1007/978-3-7908-2846-7-11

M3 - Paper

SP - 143

EP - 157

ER -