Makine Öğrenimi ile Öğrenci Başarı Tahmini: Akademik Başarıyı Etkileyen Faktörlerin Modellenmesi ve Performans Analizi

Year 2024, Volume: 2 Issue: 2, 137 - 143, 30.12.2024

Volkan Göreke

Abstract

This study aims to analyze the factors affecting students' academic success and develop prediction models based on these factors. Many features such as "Studying Hours," "Absence," and "Family Income" that affect students' performance during the education process were evaluated. In the study, student success predictions were made using machine learning models such as Random Forest, Ridge, and Deep Neural Networks and the performance of these models were compared. The importance levels of the features were calculated with the Random Forest algorithm and the most important features were determined. As a result of the analysis, the effect of removing low-importance features from the model to increase prediction performance was also examined. The findings of the study contribute to educators making data-driven decisions and developing strategies that can increase students' success.

Keywords

Artificial Intelligence, academic performance, regression

References

• [1]Demirtaş H, Güneş H. Eğitim yönetimi ve denetimi sözlüğü.Ankara: Anı Yayıncılık; 2002.
• [2]Sarıer Y. Türkiye’de öğrencilerin akademik başarısınıetkileyen faktörler bir meta-analiz çalışması. HacettepeUniversity Journal of Education. 2016; 31(3):1-1.doi:10.16986/HUJE.2016015868.
• [3]Arıkan Nİ. Akademik başarı ile seçilmiş sosyoekonomik değişkenler arasındaki ilişki üzerine bir araştırma. USBAD Uluslararası Sosyal Bilimler Akademi Dergisi. 2019;1(2):14-29.
• [4]Boyd FB. Motivation to continue: enhancing literacy learningfor struggling readers and writers. Reading and WritingQuarterly. 2002; 18:257-277.
• [5]Jackson LA, Eye A, Witt EA, Zhao Y, Fitzgerald HE. Alongitudinal study of the effects of Internet use andvideogame playing on academic performance and the rolesof gender, race, and income in these relationships.Computers in Human Behavior. 2011; 27:228-239.doi:10.1016/j.chb.2010.08.001.
• [6]Aškerc K, Gasparič RP, Bray M. Confronting the shadoweducation system: what government policies for whatprivate tutoring? C.E.P.S. Journal. 2012; 2(4).
• [7]Quigley A, MacKay-Lyons M, Eskes G. Effects of exercise oncognitive performance in older adults: a narrative review ofthe evidence, possible biological mechanisms, andrecommendations for exercise prescription. J Aging Res.2020; 2020.
• [8]Erickson KI, Voss MW, Prakash RS, Basak C, Szabo A,Chaddock L, Kramer AF. Exercise training increases size ofhippocampus and improves memory. Proc Natl Acad Sci U SA.2011; 108(7):3017-3022. doi:10.1073/pnas.1015950108.
• [9] Kim HY, LaRusso MD, Hsin LB, Harbaugh AG, Selman RL, SnowCE. Social perspective-taking performance: construct,measurement, and relations with academic performanceand engagement. J Appl Dev Psychol. 2018; 57:24-41.
• [10]Hussain S, Khan MQ. Student-performulator: predictingstudents’ academic performance at secondary andintermediate level using machine learning. Annals of DataScience. 2023; 10:637-655.
• [11]Jayaprakash S, Krishnan S, Jaiganesh V. Predicting studentsacademic performance using an improved random forestclassifier. In: 2020 International Conference on EmergingSmart Computing and Informatics (ESCI); 2020. p. 238-243.
• [12]Kumar M, Bajaj K, Sharma B, Narang S. A comparativeperformance assessment of optimized multilevel ensemblelearning model with existing classifier models. Big Data.2022; 10:371-387.
• [13]Ofori F, Maina E, Gitonga R. Using machine learningalgorithms to predict students’ performance and improvelearning outcomes: a literature-based review. J Inf Technol.2020; 4:33-55.
• [14] Asselman A, Khaldi M, Aammou S. Enhancing the predictionof student performance based on the machine learningXGBoost algorithm. Interact Learn Environ. 2023; 31:3360-3379.
• [15]Long K, Guo D, Deng L, Shen H, Zhou F, Yang Y. Cross-combination analyses of random forest feature selection and decision tree model for predicting intraoperativehypothermia in total joint arthroplasty. J Arthroplasty. 2024;1-9.
• [16]Willmott CJ, Matsuura K. Advantages of the mean absoluteerror (MAE) over the root mean square error (RMSE) inassessing average model performance. Climate Res. 2005; 30(1):79-82.
• [17]Strobl C, Boulesteix AL, Zeileis A, Hothorn T. Bias in randomforest variable importance measures: illustrations, sources,and a solution. BMC Bioinformatics. 2007; 8(1):1-21.