Prediction of the House Price Index of Turkey: A Comparative Study of Multiple Linear Regression and Artificial Neural Network Models

Download

YKE-Thesis v30.pdf

Date

2022-6-16

Author

Erdekli, Yusuf Kemal

Metadata

Show full item record

This work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License.

Item Usage Stats

319
views

407
downloads

The housing market is accepted as the triggering force for the Turkish economy. Consequently, the fluctuation in house prices is an important point of interest for both policymakers in the government and decision-makers in the construction industry. In order to be able to monitor the movements in house prices effectively, it is important to understand the relationships between the macroeconomic parameters and house price movements. The main objective of this thesis is to predict the house price index (HPI) of Turkey using macroeconomic parameters by utilizing multiple linear regression (MLR) and artificial neural network (ANN) models. Among the 25 macroeconomic parameters commonly used in previous studies reported in the literature, 9 parameters are selected as the independent variables while predicting the house price index (HPI), which is chosen as the dependent variable, and the monthly time series data of these parameters are used for the time period between January 2010 and December 2019 to train the models. Subsequently, data from January 2020 to December 2021 is used to validate the models. The relationship between these macroeconomic parameters and house price movements is examined in detail, and consumer price index (CPI), unemployment rate (UR), and brent oil price (BOP) are found as the most significant parameters in determining the house price movements in Turkey. The developed MLR and ANN models provided a high level of accuracy in learning, generalizing, and converging the time series data and produced reliable prediction values. Model validation results revealed that the ANN model developed in this thesis has a higher predictive performance than the developed MLR model and also the ANN models developed in previous studies.

Subject Keywords

House Prices, Turkey, Macroeconomic Parameters, Multiple Linear Regression (MLR), Artificial Neural Networks (ANN)

URI

https://hdl.handle.net/11511/97935

Collections

Graduate School of Natural and Applied Sciences, Thesis

Suggestions

OpenMETU
Core

Measurement of NAIRU: What is new and what is next for the Turkish economy? Us, Vuslat (Orta Doğu Teknik Üniversitesi (Ankara, Turkey), 2017-8) The Turkish economy has been exposed to several risks after the global crisis. These emanated mostly from external factors such as the adoption of unconventional monetary policies by advanced economies, the sovereign debt crisis in the Euro area and the exit from the prolonged use of extraordinary anti–crisis measures. However, policy measures adopted against these risks led to a notable slowdown in the Turkish economy, raising questions about a middle income trap. In the meantime, starting from 2012, ...
Failing promises of homeownership in Turkey Aksoy Khurami, Esma; Özdemir Sarı, Özgül Burcu; Department of City and Regional Planning (2020-12-23) Homeownership is not a new agenda of housing policy in Turkey. During the last 20 years, policies promoting homeownership have been executed, and planning has been used as a pair of tongs. Through new house building, the increase in homeownership rates is expected to provide individuals with a means of opportunity. As initial promises, homeownership is argued to provide everyone with a chance to acquire a stable home, a way of refraining from affordability problems, wealth accumulation through housing, and ...
Sequencing, pace and timing of financial liberalization process in Turkey with implications on the macroeconomic environment Ganioğlu, Aytül; Şenses, Fikret; Department of Economics (2006) This study basically analyzes timing, sequencing and pace of the financial liberalization experience of the Turkish economy in the 1980s and evaluates its implications for the crises in the Turkish economy since the 1990s. The objectives of this study are threefold: Firstly, it aims to reveal the main policy objectives and political factors pushing the government to take capital account liberalization decision in 1989. It is concluded that domestic decision makers have shaped and taken the decision of capit...
Understanding real estate bubbles: an analysis of the recent trends in the Turkish housing market Karasu, Mustafa Nusret; Cömert, Hasan; Department of Economics (2015) The main aim of this thesis is to determine whether there is a real estate bubble in Turkey recently, where the house prices are increasing rapidly. In order to do this, the thesis tries to define the bubble concept, by reviewing literature. The common view among economists is that bubbles are deviation of prices from their fundamentals which is associated with sudden increases in prices. We usually witness harsh economic crises after the bursts of bubbles in economies and this constitutes the source of the...
Empirical investigation of owner-occupiers' reinvestments in housing: the case of Ankara, Turkey Özdemir Sarı, Özgül Burcu (Springer Science and Business Media LLC, 2014-03-01) In Turkey, reinvestments in the existing housing stock are entirely dependent on households' decisions in the free market. There are no policies to consider reinvestment processes, and the body of knowledge on households' reinvestment decisions is scant. Understanding how individual reinvestment decisions are determined is vital to devise policy measures to improve the condition of the existing housing stock and neighbourhoods. In this study, an attempt is made to identify the basic motivations and factors ...

Citation Formats

Y. K. Erdekli, “Prediction of the House Price Index of Turkey: A Comparative Study of Multiple Linear Regression and Artificial Neural Network Models,” M.S. - Master of Science, Middle East Technical University, 2022.