Predicting Academic Performance with Applying Data Mining Techniques Generalizing the results of two Different Case Studies

Several research works are attempted to predict students’ academic performance and assess the evaluating students’ knowledge or detecting students’ weakness and probability of failure in final semester examinations. However, several factors affect the performance of students in different countries or even in different states of one country. Therefore, understanding these factors and analyzing the effects of each one of those factors in each country, is necessary for improving instructors’ decisions in selecting the best teaching method for helping weak students or increasing performance of other students. This study is motivated to study the students’ academic performance in high school and bachelor degree studies in Iran and comparing these analysis results with the similar study’s results in India. In this article, we used several classification and prediction algorithms for improving the accuracy of predicting students’ academic results before examination. Our experimental results show that, factors such as gender, family background, parents’ educational level, life style have important effects on students’ academic performance in both countries. We collected students’ academic performance results from their high schools and dean academic of universities. Data for this research work is collected through the interview with students and asking questions regarding their number of family members, parents’ job and their education level, family welfare. Further, with applying simple Random sampling (without replacement) method, 300 students including 100 high school and 200 college students are selected and the mental condition of these students is checked by our Psychologist for identifying their stress level or any other mentality problems.

Academic Performance, CPI (Cumulative Performance Index), Classification, Prediction, Simple Random Sampling, Decision Making.