Forecasting Dropout In University Based On Students' Background Profile Data Through Automated Machine Learning Approach
Citations Over Time
Abstract
A common research problem is predicting student dropout early and correctly based on existing education data. In recent years, machine learning has received a lot of attention in the fight against dropout. This is because machine learning technologies can successfully identify at-risk students and prepare precautionary measures in a timely manner. We consider predicting student's possible dropout rate from university programs prior to admission. To that purpose, we collect our own statistics from students who began their studies between 2014 and 2016. We are left with 2066 participants after preprocessing and cleaning. Six distinct binary classifiers, namely the Artificial Neural Network, Naive Bayes, Decision Tree, Support Vector Machine, Random Forest Tree, and k - Nearest Neighbor models, were used to predict graduations and dropouts. According to research, the average performance of six models is 84%, 80%, 77%, 82 % , 80%, and 81%. This type of research is critical in determining students' success rates at university programs based on their pre-university data.
Related Papers
- → MOOC Dropout Prediction(2017)86 cited
- → Controlled dropout: A different approach to using dropout on deep neural network(2017)38 cited
- → A model for the identification of students at risk of dropout at a university of technology(2020)18 cited
- → Using Stacked Denoising Autoencoder for the Student Dropout Prediction(2017)13 cited
- Exploration of High Rate of Dropout in Distance Educaion(2007)