Perbandingan Kinerja Naïve Bayes dengan dan Tanpa SMOTE untuk Klasifikasi Gangguan Kecemasan Mahasiswa pada Data Tidak Seimbang

Abstract

Anxiety disorders are one of the most common mental health problems experienced by university students and may affect learning concentration and academic performance. The analysis of psychological survey data using machine learning techniques can support early detection of student anxiety conditions. However, one of the main challenges in mental health data analysis is the presence of class imbalance within the dataset. This study aims to analyze the effect of applying the Synthetic Minority Oversampling Technique (SMOTE) on the performance of the Naïve Bayes algorithm for multi-class classification of student anxiety levels, which are categorized into three classes: No Stress, Eustress, and Distress. The dataset used in this research was obtained from student questionnaire data and underwent several preprocessing steps including data cleaning, feature transformation, and dataset splitting using the hold-out method with a ratio of 80% training data and 20% testing data. Model performance was evaluated using a confusion matrix with evaluation metrics including accuracy, precision, recall, and F1-score. The results show that the Naïve Bayes model without SMOTE achieved an accuracy of 0.84, precision 0.78, recall 0.41, and F1-score 0.54. After applying SMOTE, the model achieved an accuracy of 0.82, precision 0.74, recall 0.69, and F1-score 0.71. These results indicate that SMOTE improves the model's ability to detect minority classes in multi-class classification problems, although a slight decrease in overall accuracy is observed.