Implementasi Least Squares Support Vector Machine dan SMOTE untuk Klasifikasi Kesehatan Mental
Abstract
Mental health disorders such as depression, anxiety, and stress are global problems that require accurate early detection. This study proposes a mental health classification model using machine learning algorithms based on data from the Depression Anxiety Stress Scales (DASS-42) questionnaire and respondent demographic features. The main method used is Least Squares Support Vector Machine (LSSVM) combined with Synthetic Minority Oversampling Technique (SMOTE) and Backward Elimination feature selection. From testing on 39,775 respondent data, Backward Elimination successfully reduced more than 40% of the feature dimensions by selecting the most statistically significant attributes (p-value < 0.05). Oversampling with SMOTE proved successful in overcoming class imbalance in minority labels. Performance evaluation showed that LSSVM using the Radial Basis Function (RBF) kernel provided the most optimal results compared to the Linear and Polynomial kernels, with an F1-Score of 83.96% for Depression, 77.30% for Anxiety, and 82.00% for Stress. This proposed model contributes to the development of a more computational, efficient, and accurate mental health screening system.
Keywords: Backward Elimination; DASS-42; LSSVM; Mental Health; SMOTE
Abstrak
Gangguan kesehatan mental seperti depresi, kecemasan, dan stres merupakan masalah global yang memerlukan deteksi dini yang akurat. Penelitian ini mengusulkan model klasifikasi tingkat kesehatan mental menggunakan algoritma machine learning berdasarkan data kuesioner Depression Anxiety Stress Scales (DASS-42) dan fitur demografis responden. Metode utama yang digunakan adalah Least Squares Support Vector Machine (LSSVM) yang dikombinasikan dengan Synthetic Minority Oversampling Technique (SMOTE) dan seleksi fitur Backward Elimination. Dari pengujian terhadap 39.775 data responden, Backward Elimination berhasil mereduksi lebih dari 40% dimensi fitur dengan menyeleksi atribut yang paling signifikan secara statistik (p-value < 0.05). Oversampling dengan SMOTE terbukti berhasil mengatasi ketidakseimbangan kelas pada label minoritas. Evaluasi kinerja menunjukkan bahwa LSSVM menggunakan kernel Radial Basis Function (RBF) memberikan hasil paling optimal dibandingkan kernel Linear dan Polynomial, dengan pencapaian F1-Score sebesar 83.96% untuk Depresi, 77.30% untuk Kecemasan, dan 82.00% untuk Stres. Model yang diusulkan ini berkontribusi dalam pengembangan sistem screening kesehatan mental yang lebih komputasional, efisien, dan akurat.
Keywords
References
K. Vitoasmara, F. V. Hidayah, N. I. Purnamasari, R. Y. Aprillia, and L. D. D. A, “Gangguan Mental (Mental Disorders),” Student Res. J., vol. 2, no. 3, pp. 57–68, 2024.
M. A. Hakim and N. V. Aristawati, “Mengukur depresi, kecemasan, dan stres pada kelompok dewasa awal di Indonesia: Uji validitas dan reliabilitas konstruk DASS-21,” J. Psikol. Ulayat, vol. 10, no. 2, pp. 232–250, 2023, doi: 10.24854/jpu553.
N. A. SHAYAN, A.-R. NIAZI, A. M. WASEQ, and H. ÖZCEBE, “Depression, Anxiety, and Stress Scales 42 (DASS-42) in Dari-Language: Validity and Reliability Study in Adults, Herat, Afghanistan,” Bezmialem Sci., vol. 9, no. 3, pp. 356–362, 2021, doi: 10.14235/bas.galenos.2020.4250.
V. M. A. Simalango and W. F. Senjaya, “Evaluasi Algoritma Decision Tree dan Random Forest serta Efektivitas Feature Selection dalam Memprediksi Kesehatan Mental,” J. Strateg. - J. Maranatha, vol. 7, no. 1, pp. 201–215, 2025, [Online]. Available: https://strategi.it.maranatha.edu/index.php/strategi/article/view/583
S. Juwariyah, A. Hulvi, N. Riduan, and K. Kusrini, “Mengukur Faktor Demografi Psikologis: Memprediksi Depresi, Kecemasan, dan Stres dengan menggunakan Machine Learning,” Komputika J. Sist. Komput., vol. 13, no. 2, pp. 149–156, 2024, doi: 10.34010/komputika.v13i2.11793.
S. D. Amalia, M. A. Barata, and P. E. Yuwita, “Optimization of Random Forest Algorithm with Backward Elimination Method in Classification of Academic Stress Levels,” J. Appl. Informatics Comput., vol. 9, no. 3, pp. 633–641, 2025, doi: 10.30871/jaic.v9i3.9280.
E. P. Efendi, M. A. Barata, and A. D. Ardianti, “Komparasi Algoritma Svm Dan C4 . 5 Dengan Backward Elimination Untuk Klasifikasi Tingkat Stres Mahasiswa Comparative Analysis Of Svm And C4 . 5 Algorithms With Backward,” JIKO (Jurnal Inform. dan Komputer), no. 3, pp. 539–549, 2024, doi: 10.26798/jiko.v9i3.2130.
Y. Xia, J. Zhao, Q. Ding, and A. Jiang, “Incipient Chiller Fault Diagnosis Using an Optimized Least Squares Support Vector Machine With Gravitational Search Algorithm,” Front. Energy Res., vol. 9, no. November, pp. 1–13, 2021, doi: 10.3389/fenrg.2021.755649.
S. Acharya, T. Kar, U. C. Samal, and P. K. Patra, “Performance Comparison between SVM and LS-SVM for Rice Leaf Disease detection,” ICST Trans. Scalable Inf. Syst., vol. 10, no. 6, pp. 1–7, 2023, doi: 10.4108/eetsis.3940.
F. M. Basysyar, G. Dwilestari, and A. I. Purnamasari, “Analysis Student Emotions and Mental Health on Cumulative Gpa Using Machine Learning and Smote,” JITK (Jurnal Ilmu Pengetah. dan Teknol. Komputer), vol. 10, no. 2, pp. 361–368, 2024, doi: 10.33480/jitk.v10i2.5967.
M. R. Pribadi, H. D. Purnomo, and Hendry, “A three-step combination strategy for addressing outliers and class imbalance in software defect prediction,” IAES Int. J. Artif. Intell., vol. 13, no. 3, pp. 2987–2998, 2024, doi: 10.11591/ijai.v13.i3.pp2987-2998.
J. T. Hancock and T. M. Khoshgoftaar, “Survey on categorical data for neural networks,” J. Big Data, 2020, doi: 10.1186/s40537-020-00305-w.
D. H. Maulud and A. M. Abdulazeez, “A Review on Linear Regression Comprehensive in Machine Learning,” J. Appl. Sci. Technol. Trends, vol. 01, no. 04, pp. 140–147, 2020, doi: 10.38094/jastt1457.
W. Nugroho, “Optimasi Metode K-Nearest Neighbours dengan Backward Elimination Menggunakan Dataset Software Effort Estimation Bianglala Informatika,” J. Khatulistiwa Inform., vol. 8, no. 2, pp. 129–133, 2020, [Online]. Available: https://download.garuda.kemdikbud.go.id/article.php?article=2191946&val=10501&title=Optimasi Metode K-Nearest Neighbours dengan Backward Elimination Menggunakan Dataset Software Effort Estimation
Y. Desnelita, N. Nasution, L. Suryati, and F. Zoromi, “Dampak SMOTE terhadap Kinerja Random Forest Classifier berdasarkan Data Tidak seimbang Impact of SMOTE on Random Forest Classifier Performance based on Imbalanced Data,” Matrik J. Manajemen, Tek. Inform. dan Rekayasa Komput., vol. 21, no. 3, 2022, doi: 10.30812/matrik.v21i3.1726.
R. Widyana, Sumiharso, and R. M. Safitri, “Psychometric Properties of Internet-Administered Version of Depression , Anxiety and Stress Scales ( DASS-42 ) in sample Indonesian Adult Psychometric Properties of Internet-Administered Version of Depression , Anxiety and Stress Scales ( DASS-42 ) in sam,” Talent Dev Excell, no. June, 2022, [Online]. Available: https://www.researchgate.net/profile/Rahma-Widyana/publication/361179345_Psychometric_Properties_of_Internet-Administered_Version_of_Depression_Anxiety_and_Stress_Scales_DASS-42_in_sample_Indonesian_Adult/links/62a13f1d6886635d5ccaf778/Psychometric-Properties-of-Internet-Administered-Version-of-Depression-Anxiety-and-Stress-Scales-DASS-42-in-sample-Indonesian-Adult.pdf?_sg%5B0%5D=started_experiment_milestone&origin=journalDetail
N. Fathirachman Mahing, A. Lazuardi Gunawan, A. Foresta Azhar Zen, F. Abdurrachman Bachtiar, and S. Agung Wicaksono, “Klasifikasi Tingkat Stress dari Data Berbentuk Teks dengan Menggunakan Algoritma Support Vector Machine (SVM) dan Random Forest,” J. Teknol. Inf. dan Ilmu Komput., vol. 11, no. 5, pp. 1067–1076, 2024, doi: 10.25126/jtiik.1078010.
Z. Fatah and U. Hasanah, “Prediksi Tingkat Stress dan Kesehatan Mental Mahasiswa Menggunakan Algoritma SVM,” JAMASTIKA (Jurnal Mhs. Tek. Inform., vol. 4, no. 2, pp. 200–207, 2025, [Online]. Available: https://jurnal.unw.ac.id/index.php/jamastika/ article/view/4542
G. Samudra, A. W. Anggraeni, A. R. B. Jamroni, A. Sarif, and W. Wiyanto, “Efektivitas Teknik SMOTE Dalam Meningkatkan Performa Naïve Bayes Deteksi Gangguan Kecemasan Mahasiswa,” JATISI (Jurnal Tek. Inform. dan Sist. Informasi), vol. 12, no. 3, 2025, doi: 10.35957/jatisi.v12i3.12197.
T. F. Husin, M. R. Pribadi, and Yohannes, “Implementation of LSSVM in Classification of Software Defect Prediction Data with Feature Selection,” in 2022 9th International Conference on Electrical Engineering, Computer Science and Informatics (EECSI), Jakarta, 2022, pp. 126–131. doi: 10.23919/EECSI56542.2022.9946611.
How To Cite This :
Refbacks
- There are currently no refbacks.
