Predicting House Price Determinants by Using Semi Parametric Regression, Hedonic Price Model and Generalized Additive Model A Case Study of Pakistan
Author: Nauman Ahmad

The study investigate the factors affecting the house price by using hedonic price model, semi-parametric regression and machine learning technique such as Generalized Additive Model (GAM) and compares the price prediction performance with conventional parametric models. This study utilizes data set representing 6,807 single family residential home sales prices between June 2013 and July 2014 for Pakistan. Data are obtained from Pakistan Social Living standard Measurement (PSLM) and Household Integrated Economic Survey (HIES) by the Pakistan Bureau of Statistics (PBS) are incorporated to account for locational attributes of the houses. The results show that semi-Parametric regression performs better than conventional parametric model such as Hedonic price model. Generalized Additive Model outperforms than parametric and semi-parametric counterparts in both aggregate sample and disaggregate sub sample price prediction, indicating that the Generalized Additive Model (GAM) can be useful for measurement and prediction of housing sale prices. Supervisor:- Dr. Zahid Asghar

Meta Data

Keywords : Generalized Additive Model, Hedonic Price Model, Semi Parametric
Supervisor: Zahid Asghar
Cosupervisor: Amna Urooj

Related Thesis​