Comparative analysis of classification models in predicting e-learning graduation

In 2021, the number of E-Learning registrants for the Anti-Corruption Academy had decreased drastically by 52%, with the number of graduates being 71%. The Anti-Corruption Academy does not yet have a model to predict participants’ course graduation. The solution to expect someone’s graduation can use a data mining process. This study aims to predict course graduation at the Anti-Corruption Academy E-Learning by comparing classification data mining methods, namely Decision Tree, Support Vector Machine (SVM), and Random Forest, and based on accuracy values. This study predicts course graduation using the Knowledge Discovery in Databases (KDD) method on the course graduation data set and runs it on Orange software which has advantages in visual programming. The data used as attributes include gender, occupation, anti-corruption education experience, age, institution, E-Learning experience, and reasons for participating in anti-corruption E-Learning. The study’s results using these two classification methods indicate that the Random Forest method is the best in predicting the graduation course of anti-corruption academy participants with an accuracy value of 71.7%, namely proper classification. Significant factors that affect course graduation are age, domicile, a nd gender.