The Role of SMOTE in Enhancing Naive Bayes Classification for Major Choice Prediction

Abstract

This study examines the application of the Synthetic Minority Oversampling Technique (SMOTE) to address
class imbalance within a dataset used for predicting high school major selection. The dataset comprises 468 training
instances, including 306 labeled as 'IPA' and 162 labeled as 'IPS'. Despite the implementation of SMOTE, the results reveal
no significant enhancement in the predictive performance of the models, as both the SMOTE and non-SMOTE models
achieved an accuracy of 100%, an F1-score of 100%, and a recall of 100%. This finding suggests that other factors, such
as the selection of relevant features, hyperparameter tuning, and model complexity, may have a more substantial impact on
prediction performance. Additionally, the study proposes several recommendations for future research, including
conducting a more in-depth feature analysis, exploring alternative classification algorithms with advanced class imbalance
handling mechanisms, and performing meticulous hyperparameter optimization to improve overall model performance.