Comparison of Machine Learning Methods in Prediction Gini Coefficient for OECD Countries

Tuba Koç; Pelin Akın

Comparison of Machine Learning Methods in Prediction Gini Coefficient for OECD Countries

Abstract

Income inequality refers to the situation where income distribution is not shared regularly and fairly. Income inequality is among the essential problems of countries in both economic and social terms. The Gini coefficient is widely used to measure income inequality. In this study, random forest, support vector algorithms, and multiple linear regression model, which are among the machine learning algorithms, were applied to estimate the Gini coefficient of Organization for Economic Co-operation and Development (OECD) countries for 2015-2018. When the models were compared according to performance criteria, the best model was the random forest model with the highest R2 = 0.7085 and the smallest RMSE = 0.0264. According to the random forest model results, the tax revenue variable has the greatest impact on the Gini coefficient. The country with the highest Gini coefficient is Mexico, and the lowest is the Slovak Republic. Also, it has been observed that the lowest tax income value belongs to Mexico.

Keywords

Support vector, Random forest, Gini coefficient, OECD

References

[1] Niyimbanira F. Analysis of the impact of economic growth on income inequality and poverty in South Africa: the case of Mpumalanga Province. International Journal of Economics and Financial Issues, 2017.
[2] Li H, Xu L C, Zou H f. Corruption, income distribution, and growth. Economics & Politics, 2000. 12(2): p. 155-182.
[3] Peçe M A, Ceyhan M S, Akpolat A. Türkiye’de gelir dağılımının ekonomik büyümeye etkisi üzerine ekonometrik bir analiz. Uluslararası Kültürel Ve Sosyal Araştırmalar Dergisi (Uksad), 2016. 2(Special Issue 1): p. 135-148.
[4] Yazgan Ş. Kamu yatırımları dağılımının gini katsayısı ile ölçülmesi: türkiye üzerine bir uygulama (1999-2017). Uluslararası Ekonomi Siyaset İnsan ve Toplum Bilimleri Dergisi, 2018. 1(1): p. 35-44.
[5] Öz S. Gelir dağılımında gını katsayısı ve p80/p20 oranı arasındaki ilişkiler: 2000-2016 dönemi Türkiye örneği. 2019.
[6] Demir M A. Gelir dağılımı eşitsizliği ve lüks mal ithalatı arasında panel nedensellik analizi. Akademik Araştırmalar ve Çalışmalar Dergisi (AKAD), 2020. 12(23): p. 471-483.
[7] Zaman T, Dünder E, Aydın S. Gini katsayısını etkileyen faktörlerin beta regresyon yöntemi yardımı ile belirlenmesi. Erzincan Üniversitesi Fen Bilimleri Enstitüsü Dergisi, 2019. 12(1): p. 235-240.
[8] Basumatary S, Devi M, Basumatary K. Assessing the Disparities of Per-capita Electricity Consumption in North-Eastern States of India Using Gini Index and Lorenz Curve. Journal of Humanities and Social Sciences Studies, 2021. 3(1): p. 103-107.
[9] Vapnik V. Principles of risk minimization for learning theory. in Advances in neural information processing systems. 1992.
[10] Gunn S R. Support vector machines for classification and regression. ISIS technical report, 1998. 14(1): p. 5-16.

[11] Shokry A, et al., Modeling and simulation of complex nonlinear dynamic processes using data based models: Application to photo-Fenton process, in Computer Aided Chemical Engineering. 2015, Elsevier. p. 191-196.
[12] Friedman J H. Greedy function approximation: a gradient boosting machine. Annals of statistics, 2001: p. 1189-1232.
[13] Liaw A Wiener M. Classification and regression by randomForest. R news, 2002. 2(3): p. 18-22.
[14] Uğuz S, Makine Öğrenmesi Teorik Yönleri ve Python Uygulamaları ile Bir Yapay Zeka Ekolü. 2019, Ankara: Nobel. 312.
[15] Ekici M S. Vergi gelirlerini etkileyen ekonomik ve sosyal faktörler. Elektronik Sosyal Bilimler Dergisi, 2009. 8(30): p. 200-223.
[16] Balcı B İ, Alyu E. Yoksulluk ve Gelir Dağılımı Eşitsizliği: OECD ve AB Ülkeleri Panel Veri Analizi. Gaziantep Üniversitesi Sosyal Bilimler Dergisi, 2018. 17(3): p. 988-996.
[17] Yanpar A. Gelişmekte olan ülkelerde büyüme yönelimli vergi politikası. Ankara Üniversitesi, SBE-Yüksek Lisans Tezi, 2007.

Details

Primary Language

English

Subjects

Machine Learning Algorithms

Journal Section

Research Article

Authors

Tuba Koç This is me
Türkiye

Pelin Akın
Türkiye

Publication Date

July 1, 2021

Submission Date

April 5, 2021

Acceptance Date

June 6, 2021

Published in Issue

Year 2021 Volume: 4 Number: 1

IZ

https://izlik.org/JA54UT29JZ

APA

Koç, T., & Akın, P. (2021). Comparison of Machine Learning Methods in Prediction Gini Coefficient for OECD Countries. Data Science and Applications, 4(1), 16-21. https://izlik.org/JA54UT29JZ

AMA

1.Koç T, Akın P. Comparison of Machine Learning Methods in Prediction Gini Coefficient for OECD Countries. DataSCI. 2021;4(1):16-21. https://izlik.org/JA54UT29JZ

Chicago

Koç, Tuba, and Pelin Akın. 2021. “Comparison of Machine Learning Methods in Prediction Gini Coefficient for OECD Countries”. Data Science and Applications 4 (1): 16-21. https://izlik.org/JA54UT29JZ.

EndNote

Koç T, Akın P (July 1, 2021) Comparison of Machine Learning Methods in Prediction Gini Coefficient for OECD Countries. Data Science and Applications 4 1 16–21.

IEEE

[1]T. Koç and P. Akın, “Comparison of Machine Learning Methods in Prediction Gini Coefficient for OECD Countries”, DataSCI, vol. 4, no. 1, pp. 16–21, July 2021, [Online]. Available: https://izlik.org/JA54UT29JZ

ISNAD

Koç, Tuba - Akın, Pelin. “Comparison of Machine Learning Methods in Prediction Gini Coefficient for OECD Countries”. Data Science and Applications 4/1 (July 1, 2021): 16-21. https://izlik.org/JA54UT29JZ.

JAMA

1.Koç T, Akın P. Comparison of Machine Learning Methods in Prediction Gini Coefficient for OECD Countries. DataSCI. 2021;4:16–21.

MLA

Koç, Tuba, and Pelin Akın. “Comparison of Machine Learning Methods in Prediction Gini Coefficient for OECD Countries”. Data Science and Applications, vol. 4, no. 1, July 2021, pp. 16-21, https://izlik.org/JA54UT29JZ.

Vancouver

1.Tuba Koç, Pelin Akın. Comparison of Machine Learning Methods in Prediction Gini Coefficient for OECD Countries. DataSCI [Internet]. 2021 Jul. 1;4(1):16-21. Available from: https://izlik.org/JA54UT29JZ