Read, Simon (2011) A Bayesian approach to the Bernoulli spatial scan statistic. Working Paper.
Abstract
This document describes a novel approach to finding localised clusters in spatially distributed, binary labelled point data. The frequentist spatial scan statistic, introduced by Martin Kulldorff in 1995, was developed into a Bayesian spatial scan statistic for areal data by Daniel Neill, circa 2006, where computationally expensive Monte Carlo testing is replaced by the use of historical data and expert judgement. Following Neill's approach, I present here my derivation of a Bayesian spatial scan statistic for binary labelled point data. I have also developed a method for replacing historic data with expert judgement, by using a prior probability distribution of relative risk. Please note this document describes work in progress, and content may be subject to revision.
Metadata
Item Type: | Monograph |
---|---|
Authors/Creators: |
|
Keywords: | Bayesian Bernoulli spatial scan statistic Omega |
Dates: |
|
Institution: | The University of Sheffield |
Academic Units: | The University of Sheffield > Faculty of Social Sciences (Sheffield) > Information School (Sheffield) |
Depositing User: | Dr Simon Read |
Date Deposited: | 21 Sep 2011 16:31 |
Last Modified: | 05 Jun 2014 08:25 |
Status: | In preparation |
Open Archives Initiative ID (OAI ID): | oai:eprints.whiterose.ac.uk:43245 |
Download
Filename: BayesianBernoulliSSS.pdf
Description: Working paper outlining a novel approach to the Bernoulli spatial scan statistic