Analysing Household Food Consumption in Turkey Using Machine Learning Techniques

• Published in: AGRIS on-line Papers in Economics and Informatics Vol. 16; no. 2; pp. 97 - 105
• Main Authors: Oztornaci, Burak, Ata, Baris, Kartal, Serkan
• Format: Journal Article
• Language: English
• Published: Prague Faculty of Economics and Management CULS Prague 01.06.2024
• Subjects: Algorithms; Consumer Price Index; Data analysis; Datasets; Decision trees; Econometrics; Expenditures; Family income; Food; Food consumption; Food prices; Households; Impact analysis; Machine learning; Methods; Multilayers; Regression analysis; Statistical analysis; Support vector machines; Variables; Turkey
• ISSN: 1804-1930; 1804-1930
• DOI: 10.7160/aol.2024.160207

The fluctuations in food prices have highlighted the significance of analysing the factors influencing household food consumption. Recent advancements in data analysis have opened new avenues for investigating this subject. While studies have employed novel data analysis methods to examine the factors impacting household food consumption, the effect of the chosen analysis method on the research outcome remains unexplored. In this study, we aimed to investigate household food consumption in Turkey between 2012-2019 using various data analysis techniques (Linear Regression, Support Vector Machine, Random Forest, eXtreme Gradient Boosting, and Multi-Layer Perception). Our findings reveal that income emerged as the most influential factor in household food consumption across all methods. However, the impact of other factors varied depending on the method employed. This suggests that the method chosen to analyse factors other than income in studies of this nature can significantly impact the results. Researchers should exercise caution when selecting their analysis method.