This repository contains data mining on the information of whole Iranian students. I examined students' information and obtained reasons for academic growth, academic failure, and dropout.
I investigated the critical levels of dropping out of school and factors which have effect on it. Consequently, you can see the results in LeavingSchool.ipynb file. To cite an example, GPA has the most effect on dropping out of school. The following image indicates the relation between students' GPA and dropping out of school.
By considering the image, NME means need more effect, ACC stands for acceptable, GOO means good and VGO means very good.
You can also find the relation between expelled from school and other factors such as students' GPA, father's job etc. in FiredFromSchool.ipynb file. In Addition, in this file it has been shown which factors have the most effect on being expelled from school.
In both BalanceAveragePredict_Grade9.ipynb and BalanceAveragePredict_Adulthood.ipynb I reported the results of GPA prediction for grade 9 students and adult students, respectively. Therefore, in GPA prediction for grade 9 students, the best prediction accuracy was achieved by Naïve Bayes which is 0.656 with 0.653 F1 score.