Comparison of OLS Regression and Robust Regression in Overcoming Outlier Problems (Case Study: Cost of Living Data for Urban Areas in Indonesia)
DOI:
https://doi.org/10.30829/zero.v8i2.21345Keywords:
Robust Geographically, Weighted Regression, Stunting, and ModelingAbstract
Multiple regression analysis in quantitative statistical studies describes the relationship between independent and dependent variables. On the other hand, outliers in a set of data can have an unfavourable influence on data analysis, such as high residuals, significant variances, and bias, and can even cause errors in decision-making. It can be done in several ways to overcome the outlier problem in multiple linear regression analysis, including using robust regression or Ordinary Least Square (OLS) Regression by removing data indicated as an outlier first. The OLS Regression method forms a regression model by minimizing the sum of squared residuals from the estimator of the regression equation. Meanwhile, robust regression is closer to the average parameters and variance-covariance of a particular estimator, namely by standardizing the estimator for the average parameters and variance-covariance in such a way as to produce a consistent estimator for these parameters. This research aims to compare the OLS Regression and robust regression methods as alternatives for dealing with outlier problems in data. The data used in this research is secondary data (cost of living) from the Cost of Living Survey conducted by The Central Statistics Agency of the Republic of Indonesia in 2018. The stages of this research method are literature study, data collection, descriptive analysis to see the characteristics of the data, forming a regression model using the OLS Regression method, testing classical assumptions, creating a new regression model OLS Regression, forming a regression model with Robust Regression, calculating the MSE (Mean Square Error) of each regression model formed, determining the best regression model, The results of the research show that for the cost of living data, the best regression model is obtained through the OLS Regression method with data without outliers, namely .References
Afifah, R., Andriyana, Y., & Jaya, I. G. N. M. (2017). Robust geographically weighted regression with least absolute deviation method in case of poverty in Java Island. AIP Conference Proceedings, 1827. https://doi.org/10.1063/1.4979439
Arum, P. R., Ridwan, M., Alfidayanti, I., & Wasono, R. (2024). ROBUST GEOGRAPHICALLY WEIGHTED REGRESSION WITH LEAST ABSOLUTE DEVIATION (LAD) ESTIMATION AND M-ESTIMATION ON GRDP OF WEST JAVA PROVINCE. BAREKENG: Jurnal Ilmu Matematika Dan Terapan, 18(3), 1573-1584. https://doi.org/10.30598/barekengvol18iss3pp1573-1584
Bekti, R. D., Nurhadiyanti, G., & Irwansyah, E. (2014). Spatial pattern of diarrhea based on regional economic and environment by spatial autoregressive model. 1621(November), 454-461. https://doi.org/10.1063/1.4898506
Chang, P. C., Wang, Y. W., & Liu, C. H. (2007). The development of a weighted evolving fuzzy neural network for PCB sales forecasting. Expert Systems with Applications, 32(1), 86-96. https://doi.org/10.1016/j.eswa.2005.11.021
de Myttenaere, A., Golden, B., Le Grand, B., & Rossi, F. (2016). Mean Absolute Percentage Error for regression models. Neurocomputing, 192, 38-48. https://doi.org/10.1016/j.neucom.2015.12.114
Fitrawaty, Mardiyah, A., & Maipita, I. (2020). The Determination Analysis of Inflation in North Sumatra. 124, 375-383. https://doi.org/10.2991/aebmr.k.200305.093
Hezaveh, A. M., Arvin, R., & Cherry, C. R. (2019). A geographically weighted regression to estimate the comprehensive cost of traffic crashes at a zonal level. Accident Analysis and Prevention, 131, 15-24. https://doi.org/10.1016/j.aap.2019.05.028
Oshan, T. M., Smith, J. P., & Fotheringham, A. S. (2020). Targeting the spatial context of obesity determinants via multiscale geographically weighted regression. International Journal of Health Geographics, 19(1). https://doi.org/10.1186/s12942-020-00204-6
O'Sullivan, D. (2003). Geographically Weighted Regression: The Analysis of Spatially Varying Relationships (review). Geographical Analysis, 35(3), 272-275. https://doi.org/10.1353/geo.2003.0008
Partner, A., & Vernitski, A. (2023). Maths lecturers in denial about their own maths practice? A case of teaching matrix operations to undergraduate students. MSOR Connections, 21(3), 18-28. https://doi.org/10.21100/msor.v21i3.1353
Rahayu, L., Ulfa, E. M., Sasmita, N. R., Sofyan, H., Kruba, R., Mardalena, S., & Saputra, A. (2023). Unraveling Geospatial Determinants: Robust Geographically Weighted Regression Analysis of Maternal Mortality in Indonesia. Infolitika Journal of Data Science, 1(2), 73-81. https://doi.org/10.60084/ijds.v1i2.133
Salsabila Rahman, A., Maria Tinungki, G., & Tri Herdiani, E. (2023). Robust Geographically Weighted Regression Model on Poverty Data in South Sulawesi in 2019. International Journal of Research Publications, 131(1). https://doi.org/10.47119/ijrp1001311820235415
Saralajew, S., Holdijk, L., Rees, M., & Villmann, T. (2018). Prototype-based Neural Network Layers: Incorporating Vector Quantization. 1-19.
Sugasawa, S., & Murakami, D. (2022). Adaptively Robust Geographically Weighted Regression.
Wang, M., Zhu, C., Wang, F., Li, T., & Zhang, X. (2020). Multi-factor of path planning based on an ant colony optimization algorithm. Annals of GIS, 26(2), 101-112. https://doi.org/10.1080/19475683.2020.1755725
Wulandari, P. P., Djuraidah, A., & Wigena, A. H. (2019). Robust Geographically Weighted Regression Modeling using Least Absolute Deviation and M-Estimator. International Journal of Scientific Research in Science, Engineering and Technology, 238-245. https://doi.org/10.32628/ijsrset196123
Zaki, U. H. H., Ibrahim, R., Halim, S. A., & Kamsani, I. I. (2024). Prioritized Text Detergent: Comparing Two Judgment Scales of Analytic Hierarchy Process on Prioritizing Pre-Processing Techniques on Social Media Sentiment Analysis. Baghdad Science Journal, 21(2), 662-683. https://doi.org/10.21123/bsj.2024.9750
Downloads
Published
Issue
Section
License
Authors who publish with this journal agree to the following terms:
- Authors retain copyright and grant the journal right of first publication with the work simultaneously licensed under a Creative Commons Attribution-ShareAlike 4.0 International License that allows others to share the work with an acknowledgement of the work's authorship and initial publication in this journal.
- Authors are able to enter into separate, additional contractual arrangements for the non-exclusive distribution of the journal's published version of the work (e.g., post it to an institutional repository or publish it in a book), with an acknowledgement of its initial publication in this journal.
- Authors are permitted and encouraged to post their work online (e.g., in institutional repositories or on their website) prior to and during the submission process, as it can lead to productive exchanges, as well as earlier and greater citation of published work (See The Effect of Open Access).
