Master's Theses

Department

Geosciences

Degree Name

Master of Science (MS)

Abstract

The purpose of this research is to examine the functionality in utilizing Random Forest Regression (RFR) Variable Importance (VI) values in characterizing neighborhoods based on the attributes of existing housing units by creating an automated GIS tool. An important concept that has been implemented in the past in real-estate valuation is the concept of Hedonic Price Modeling (HPM), which uses regression techniques to identify the impacts that individual attributes have on the cost of a good in a heterogenous market outside of mere utility. The benefit of this research is to produce a tool that automates the RFR process such that city planners and GIS analysis with access to ArcGIS Pro software have the capability of identifying neighborhoods that characterize specific housing value ranges with real-world examples utilizing multiple data types. From this research it was found that VI is a valid method for visualizing characteristic neighborhoods based on the housing attributes for values within a specific range, but in terms of spatial analysis other methods need to be implemented into the analysis other the VI factors.

Keywords

ArcGIS, Random Forest Regression, Python Scripting, Characteristic Neighborhoods, Variable Importance

Advisor

Dr. Richard Lisichenko

Date of Award

Spring 2023

Document Type

Thesis

Rights

© The Author(s)

Comments

For questions contact ScholarsRepository@fhsu.edu

Download


Share

COinS