Estimating LGS Revolution History and Kemalism subtest correct numbers using artificial intelligence methods

In this study, it is aimed to estimate the correct numbers in the High School Transition System (LGS) Revolution History and Kemalism subtest according to the exam grades of the secondary school Social Studies course and the History of Revolution and Kemalism course. Besides, it is aimed to compare the performances of multiple linear regression and data mining algorithms at predictive levels. The research was carried out according to the correlational survey model, one of the quantitative research methods. The universe of the research consists of students who study in public secondary schools affiliated to the Ministry of National Education in Kütahya and enter LGS in the 8th grade in the 2017-2018 academic year. Convenient sampling method was used in the study. As a result of the research, it is seen that there is a statistically significant, positive moderate relationship between all variables and the number of correct scores. According to the multiple linear regression analysis, the R2 value is 52% and the MSE value is .0242. Within the scope of other analyzes; in the results of general linear regression, regression trees, gaussian process regression, support vector machine, ensemble machine learning, deep learning, random forest and artificial neural network analysis, it is seen that MSE values vary between .024 and .044, and R2 values between .12 and .52. When the results obtained are compared with the results of the multiple linear regression analysis performed with the classical method, it is observed that the MSE performance values is higher than the MSE performance values obtained from the multiple linear regression analysis. After the comparison of MSE performances, it was examined whether the performance of artificial neural network analysis, which was claimed to be more successful than other analysis methods, but underperformed in this study, could be manually increased. As a result, according to the modified artificial neural network results, the R2 value is 62% and the MSE value is .019. Keywords: Artificial intelligence, multiple linear regression, neural network, social studies.