Dissertation/Thesis Abstract

Local spatial modeling using Geographically Weighted Regression (GWR)
by Wankie, Che, M.S., California State University, Long Beach, 2013, 112; 1523087
Abstract (Summary)

Linear regression modeling is a technique used in several fields for modeling, analyzing, and predicting the relationships between variables. In spatial analysis, regression modeling has been used extensively as a general technique to analyze geographic variables. For example, in spatial epidemiology, this technique has been used to study the spatial distribution of diseases. This thesis explored a regression technique, Geographically Weighted Regression (GWR), which unlike linear regression (ordinary least squares) accounts for spatial non-stationarity. GWR is a way of exploring spatial non-stationarity by calibrating a multiple regression model which allows different relationships to exist at different points in space. This thesis also explored the technique of GWR as a "proper" statistical model and how estimates may be obtained. Local entropy map, a statistical technique in spatial data mining, was explored to examine spatial patterns based on the bivariate relationships between the dependent and each explanatory variable. Data were analyzed to explore the performance of GWR in comparison to OLS and to examine multivariate relationships at local regions using local entropy map.

Indexing (document details)
Advisor: Safer, Alan
Commitee: Korosteleva, Olga, Suaray, Kagba
School: California State University, Long Beach
Department: Applied Statistics
School Location: United States -- California
Source: MAI 52/01M(E), Masters Abstracts International
Source Type: DISSERTATION
Subjects: Geographic information science, Statistics
Keywords:
Publication Number: 1523087
ISBN: 978-1-303-20291-9
Copyright © 2019 ProQuest LLC. All rights reserved. Terms and Conditions Privacy Policy Cookie Policy
ProQuest