APPLYING MACHINE LEARNING TECHNIQUES IN PROCESSING STUDENT DATA TO ASSIST UNIVERSITY ADMISSION

Abstract

The article presents research results on building a student data set and the results of applying machine learning techniques to make a prediction program for the type of student graduation, predicting factors in the student enrollment mix that affects student learning outcomes. To solve the above problems, we conducted a research on the primary education data set of Hanoi Metropolitan University (data for 5 years from 2016 to 2020). Machine learning techniques used include Logistic Regression technique (to predict student graduation results) and an improved technique of Linear discriminant analysis technique (to predict important factors affecting student learning outcomes) - Discriminative Feature Selection technique. From there, the authors make recommendations on the trend of enrollment at university level, some recommendations on
training organization and enrollment strategy for Hanoi Metropolitan University.