Investigating House Price Predictors: A Random Forest Application
(2023) NEKN01 20231Department of Economics
- Abstract
- In macroeconomics, the housing market has been well-researched due to its significance for economic and financial stability. This study investigated the predictive importance of selected fundamental and non-fundamental determinants of house price in three different markets: the United States, Sweden, and Japan, across three different periods. By applying a random forest model, this thesis attempted to contribute to the growing research in interpretable machine learning applications in the field of economics. The findings presented notable differences in the predictors that cause house price growth fluctuations across the countries analyzed, implying differences in the underlying market and economic conditions. While real and sentiment... (More)
- In macroeconomics, the housing market has been well-researched due to its significance for economic and financial stability. This study investigated the predictive importance of selected fundamental and non-fundamental determinants of house price in three different markets: the United States, Sweden, and Japan, across three different periods. By applying a random forest model, this thesis attempted to contribute to the growing research in interpretable machine learning applications in the field of economics. The findings presented notable differences in the predictors that cause house price growth fluctuations across the countries analyzed, implying differences in the underlying market and economic conditions. While real and sentiment indicators were more pronounced for Sweden and Japan, real variables were more significant for US house price growth compared to other variables. Although the results can be tied to the economic state corresponding to the period, the drawbacks of the methodology deterred definitive conclusions on the direction of the predictor's impact on house price growth. (Less)
Please use this url to cite or link to this publication:
http://lup.lub.lu.se/student-papers/record/9130550
- author
- Imaha, Aminath LU
- supervisor
- organization
- course
- NEKN01 20231
- year
- 2023
- type
- H2 - Master's Degree (Two Years)
- subject
- keywords
- housing, predictive importance, random forests
- language
- English
- id
- 9130550
- date added to LUP
- 2023-09-12 15:38:09
- date last changed
- 2023-09-12 15:38:09
@misc{9130550, abstract = {{In macroeconomics, the housing market has been well-researched due to its significance for economic and financial stability. This study investigated the predictive importance of selected fundamental and non-fundamental determinants of house price in three different markets: the United States, Sweden, and Japan, across three different periods. By applying a random forest model, this thesis attempted to contribute to the growing research in interpretable machine learning applications in the field of economics. The findings presented notable differences in the predictors that cause house price growth fluctuations across the countries analyzed, implying differences in the underlying market and economic conditions. While real and sentiment indicators were more pronounced for Sweden and Japan, real variables were more significant for US house price growth compared to other variables. Although the results can be tied to the economic state corresponding to the period, the drawbacks of the methodology deterred definitive conclusions on the direction of the predictor's impact on house price growth.}}, author = {{Imaha, Aminath}}, language = {{eng}}, note = {{Student Paper}}, title = {{Investigating House Price Predictors: A Random Forest Application}}, year = {{2023}}, }