ابوالفضل عبدلي آراني

Today, a large amount of educational data is produced and stored in schools and universities. To identify and predict the situation of students, analyzing this information is considered necessary and important. Considering the importance of students' performance and their eva‎luation, it can have a significant impact on improving the level of training and the performance of workshops. Therefore, in this research, we intend to help improve the level of training by using data mining to predict the performance of students in the workshop In computer science, extracting knowledge from the database is called data mining, which causes new and useful knowledge to be obtained from the huge amount of stored information. It is also educational data mining, understanding students' behavior and discovering and identifying patterns hidden in them to advise students. The statistical population of this research is about 300 samples of undergraduate students' workshops of the University of Science and Technology, which were measured and collected by a digital caliper. This data includes one of the examples of students' pieces that are made in the workshop of the University of Science and Technology. The characteristics of the data include the characteristics of the conical part, which are identified by the expert. In this research, a comprehensive statistical analysis of the characteristics under study and the existing patterns among them has been discussed. In predicting the gender of the students, due to the imbalance between the data, which leads to poor predictions, therefore, the smooth algorithm was used to balance the data. In order to validate the performance of the implemented models, this research has used the cross-validation method from one to ten subsets. Also, in order to predict students' grades, cumulative models and types of reinforcement gradient models have been used, and statistical analysis of students' grades and their genetics has been done using the data of Golestan Science and Technology.

داده كاوي , ، تحليل آماري , انتخاب ويژگي , داده هاي اموزشي

data mining , statistical analysis , , feature selection , training data

Abolfazl Abdoli Arani

Dr. Mohammad Reza Mohammad Aliha