Date of Graduation

5-2024

Document Type

Thesis

Degree Name

Master of Science in Statistics and Analytics (MS)

Degree Level

Graduate

Department

Statistics and Analytics

Advisor/Mentor

Robinson, Samantha E.

Committee Member

Petris, Giovanni G.

Second Committee Member

Keiffer, Elizabeth A.

Keywords

Appraisals; Hedonics; House Prices

Abstract

Multiple hedonic models and an automatic appraiser model were used to create a residential house’s estimated sales price. The goal is to use the limited data available to a REALTOR® to estimate the future sales price of a residential home without the aid of pictures of the property or viewing the physical property. The first model automates some of the actions of an appraiser by finding comparable sales based on proximity, based both on distance between houses and characteristics of the houses, and then calculating a weighted average price for an estimated sales price of future sales. If the model cannot find enough suitable comparison properties, the property does not get an estimate from this model and is labeled atypical. Also, a combination of models was used. The appraiser model was used to identify houses that are typical and atypical. Multiple linear regression, Ridge regression, Lasso regression, log-linear, and Box-Cox were all used on the data set to see which performed better on the full data set and reduced data sets based on the automatic model. Root mean squared error (RMSE) was used to compare the performance of the different models. Box-Cox performed the best on the full data set where the middle 50% of residuals were between -$12,982 and $18,403 with a RMSE of $61,002.46. On the typical data set a linear model performed the best but had a middle 50% of residuals between -$15,055 and $18,886 while reducing the RMSE to $41,327.09. Box-Cox performed the best on the atypical data set where the middle 50% of residuals were between -$45,252 and $27,513 with a RMSE of $112,852. While the residuals are still too large to make this an acceptable model in place of an appraiser, the coefficients give insight to the hedonic nature of what buyers want. Square footage, number of bathrooms, and number of garages all had a positive impact on price. The number of bedrooms had a negative impact on sales price while holding the other variables constant, which is initially a unintuitive result.

Download

Share

COinS