Parameter estimation for the stochastic SIS epidemic model

Research output: Contribution to conferenceSpeech

Abstract

In this work [1] we estimate the parameters in the stochastic SIS (susceptible-infected-susceptible) epidemic model [2] by using pseudo-maximum likelihood estimation (pseudo-MLE) and least squares estimation. We obtain the point estimators and 100(1−α)% confidence intervals as well as 100(1 − α)% joint confidence regions by applying least squares techniques. The pseudo-MLEs have almost the same form as the least squares case. We also obtain the exact as well as the asymptotic 100(1 − α)% joint confidence regions for the pseudo-MLEs. Computer simulations are used to illustrate our theory.
Original languageEnglish
Publication statusUnpublished - Jun 2015
Event26th Numerical Analysis Conference - University of Strathclyde, Glasgow, United Kingdom
Duration: 23 Jun 201526 Jun 2015

Conference

Conference26th Numerical Analysis Conference
CountryUnited Kingdom
CityGlasgow
Period23/06/1526/06/15

Fingerprint

Confidence Region
Epidemic Model
Least Squares
Parameter Estimation
Pseudo-maximum Likelihood
Least Squares Estimation
Maximum Likelihood Estimation
Confidence interval
Computer Simulation
Estimator
Estimate
Form

Keywords

  • susceptible-infected-susceptible model
  • stochastic SIS
  • parameter estimation

Cite this

Gray, A., Pan, J., Greenhalgh, D., & Mao, X. (2015). Parameter estimation for the stochastic SIS epidemic model. 26th Numerical Analysis Conference, Glasgow, United Kingdom.
@conference{23df4b0f66cf44eaaa42687fb603d7d4,
title = "Parameter estimation for the stochastic SIS epidemic model",
abstract = "In this work [1] we estimate the parameters in the stochastic SIS (susceptible-infected-susceptible) epidemic model [2] by using pseudo-maximum likelihood estimation (pseudo-MLE) and least squares estimation. We obtain the point estimators and 100(1−α){\%} confidence intervals as well as 100(1 − α){\%} joint confidence regions by applying least squares techniques. The pseudo-MLEs have almost the same form as the least squares case. We also obtain the exact as well as the asymptotic 100(1 − α){\%} joint confidence regions for the pseudo-MLEs. Computer simulations are used to illustrate our theory.",
keywords = "susceptible-infected-susceptible model, stochastic SIS, parameter estimation",
author = "Alison Gray and Jiafeng Pan and David Greenhalgh and Xuerong Mao",
year = "2015",
month = "6",
language = "English",
note = "26th Numerical Analysis Conference ; Conference date: 23-06-2015 Through 26-06-2015",

}

Gray, A, Pan, J, Greenhalgh, D & Mao, X 2015, 'Parameter estimation for the stochastic SIS epidemic model', 26th Numerical Analysis Conference, Glasgow, United Kingdom, 23/06/15 - 26/06/15.

Parameter estimation for the stochastic SIS epidemic model. / Gray, Alison; Pan, Jiafeng; Greenhalgh, David; Mao, Xuerong.

2015. 26th Numerical Analysis Conference, Glasgow, United Kingdom.

Research output: Contribution to conferenceSpeech

TY - CONF

T1 - Parameter estimation for the stochastic SIS epidemic model

AU - Gray, Alison

AU - Pan, Jiafeng

AU - Greenhalgh, David

AU - Mao, Xuerong

PY - 2015/6

Y1 - 2015/6

N2 - In this work [1] we estimate the parameters in the stochastic SIS (susceptible-infected-susceptible) epidemic model [2] by using pseudo-maximum likelihood estimation (pseudo-MLE) and least squares estimation. We obtain the point estimators and 100(1−α)% confidence intervals as well as 100(1 − α)% joint confidence regions by applying least squares techniques. The pseudo-MLEs have almost the same form as the least squares case. We also obtain the exact as well as the asymptotic 100(1 − α)% joint confidence regions for the pseudo-MLEs. Computer simulations are used to illustrate our theory.

AB - In this work [1] we estimate the parameters in the stochastic SIS (susceptible-infected-susceptible) epidemic model [2] by using pseudo-maximum likelihood estimation (pseudo-MLE) and least squares estimation. We obtain the point estimators and 100(1−α)% confidence intervals as well as 100(1 − α)% joint confidence regions by applying least squares techniques. The pseudo-MLEs have almost the same form as the least squares case. We also obtain the exact as well as the asymptotic 100(1 − α)% joint confidence regions for the pseudo-MLEs. Computer simulations are used to illustrate our theory.

KW - susceptible-infected-susceptible model

KW - stochastic SIS

KW - parameter estimation

UR - http://numericalanalysisconference.org.uk/

M3 - Speech

ER -

Gray A, Pan J, Greenhalgh D, Mao X. Parameter estimation for the stochastic SIS epidemic model. 2015. 26th Numerical Analysis Conference, Glasgow, United Kingdom.