Klasifikasi Tingkat Kecemasan pada Mahasiswa Fakultas Sains dan Analitika Data ITS Menggunakan Fuzzy Support Vector Machine

Anggraeni, Ayu (2025) Klasifikasi Tingkat Kecemasan pada Mahasiswa Fakultas Sains dan Analitika Data ITS Menggunakan Fuzzy Support Vector Machine. Other thesis, Institut Teknologi Sepuluh Nopember.

Text 5003211053-Undergraduate_Thesis.pdf - Accepted Version Restricted to Repository staff only Download (6MB) | Request a copy

Abstract

Penelitian ini mengusulkan sebuah metode untuk mengklasifikasikan tingkat kecemasan yang dialami oleh mahasiswa sarjana Fakultas Sains dan Analitika Data Institut Teknologi Sepuluh Nopember dengan menggunakan Fuzzy Support Vector Machine (FSVM). Keunggulan dari metode ini yaitu kemampuannya dalam menangani data imbalanced dengan cara memberikan bobot keanggotaan fuzzy sehingga setiap sampel memiliki kontribusi yang berbeda terhadap pembentukan hyperplane. Data dikumpulkan melalui survei menggunakan Depression Anxiety Stress Scale-21 (DASS-21), salah satu kuesioner yang mengukur tingkat kecemasan. Skor total kecemasan dari DASS-21 yang terdiri dari 5 kategori disederhanakan menjadi dua kategori, yaitu normal dan cemas. Model FSVM dilatih dengan kernel kategorik dan stratified 5-fold cross validation lalu dievaluasi menggunakan confusion matrix dan nilai AUC. Model dibandingkan dengan metode klasifikasi lain seperti SVM dan SVM dengan random oversampling. Hasil penelitian menunjukkan bahwa klasifikasi menggunakan FSVM dengan bobot fuzzy 0,2 untuk kelas mayoritas memberikan performa paling optimum dibandingkan metode lainnya. Model terbaik diperoleh pada saat fold ke-1 dengan nilai akurasi, sensitivity, specifity, dan AUC secara berturut-turut 0,600; 0,519; 1,000; dan 0,759. Sementara itu, penggunaan model terbaik untuk prediksi data testing menghasilkan akurasi, sensitivity, specifity, dan AUC secara berturut-turut 0,679; 0,691; 0,615; dan 0,700. Selain itu, hasil feature importance menunjukkan bahwa variabel yang berpengaruh besar terhadap model prediksi yaitu nyeri lambung dengan penurunan nilai AUC sebesar 0,107.
========================================================================================================================================
This study proposes a method to classify the anxiety levels experienced by undergraduate students of the Faculty of Science and Data Analytics Institut Teknologi Sepuluh Nopember using the Fuzzy Support Vector Machine. The main advantage of this method lies in its ability to handle imbalanced data by assigning fuzzy membership weights, allowing each sample to contribute differently to the formation of the hyperplane. Data were collected through a survey using the Depression Anxiety Stress Scale-21 (DASS-21), a widely used questionnaire to measure anxiety levels. The total anxiety scores from DASS-21, originally consisting of five categories, were simplified into two categories: normal and anxious. The FSVM model was trained using a categorical kernel and stratified 5-fold cross validation then evaluated using confusion matrix and the Area Under the Curve (AUC) metric. It was then compared with other classification methods, including standard SVM and SVM with random oversampling. The results of this study indicate that classification using FSVM with a fuzzy weight of 0.2 for the majority class yields the most optimal performance among the methods evaluated. The best model performance was obtained in fold 1, with accuracy, sensitivity, specificity, and AUC scores of 0.600; 0.519; 1.000; and 0.759, respectively. When applied to the testing data, the best model achieved accuracy, sensitivity, specificity, and AUC scores of 0.679; 0.691; 0.615; and 0.700, respectively. Additionally, the feature importance analysis indicated that the most influential variable for the prediction model was stomach pain, as evidenced by a decrease in AUC of 0.107 when this variable was removed.

Item Type:	Thesis (Other)
Uncontrolled Keywords:	DASS-21, Data Imbalanced, Fuzzy Support Vector Machine, Gangguan Kecemasan, Klasifikasi, Anxiety Disorder, Classification, DASS-21, Fuzzy Support Vector Machine, Imbalanced Data
Subjects:	Q Science > QA Mathematics > QA248_Fuzzy Sets
Divisions:	Faculty of Civil Engineering and Planning > Civil Engineering > 22301-(D4) Diploma 4
Depositing User:	Ayu Anggraeni
Date Deposited:	01 Aug 2025 07:35
Last Modified:	01 Aug 2025 07:35
URI:	http://repository.its.ac.id/id/eprint/125682

Actions (login required)

View Item