Drawing on twelve years of administrative data at a large public university in the us, we find that dropout prediction at the end of the second year has a 20% higher auc than at the time of enrollment in a random forest model. Mooc dropout prediction using machine learning techniques The conference is organised annually by the society f
To achieve this, we study four datasets from three courses, and compare the performance of two approaches for building models for predicting who are likely to drop out. We report associations between our prediction and academic outcomes, prompting scrutiny of discrepancies between credit hour designation and course load prediction at the course level. We proposed a student's dropout prediction model using an intuitionistic fuzzy set and an xgboost algorithm called stou2pm