SHAP dalam Explainable AI untuk Deteksi Depresi Mahasiswa Menggunakan AdaBoost
Abstract
Depresi pada mahasiswa merupakan masalah kesehatan mental yang perlu dideteksi secara dini agar intervensi dapat dilakukan dengan lebih tepat. Penelitian ini berfokus pada deteksi depresi mahasiswa menggunakan algoritma Adaptive Boosting (AdaBoost) yang dipadukan dengan metode Explainable Artifical Intelligence (XAI) berupa SHAP untuk memberikan interpretasi terhadap kontribusi setiap fitur dalam proses prediksi. Dua dataset depresi mahasiswa dari Kaggle digunakan untuk membandingkan pengaruh kualitas dan karakteristik data terhadap performa model. Pada masing-masing dataset, dilakukan pelatihan model AdaBoost dan evaluasi menggunakan metrik akurasi, presisi, recall, dan F1-score, serta analisis interpretabilitas menggunakan SHAP yang kemudian dibandingkan dengan Feature Importance bawaan AdaBoost. Hasil pengujian menunjukkan bahwa dataset Jatmiko et al. menghasilkan akurasi yang lebih tinggi dibandingkan dataset Shodolamu Opeyami, sehingga diniliai memiliki kualitas yang lebih baik. Analisis SHAP juga mengungkap bahwa fitur terkait pikiran bunuh diri merupakan faktor paling dominan dalam prediksi pada kedua dataset. Selain itu, SHAP terbukti lebih stabil dan reliabel dibandingkan Feature Importance, sehingga lebih sesuai digunakan sebagai metode interpretabilitas pada model deteksi depresi mahasiswa.
Kata kunci: depresi mahasiswa, machine learnng, adaboost, explainable ai, shap, interpretabilitas
References
S. Pattnaik, A. Nayak, S. Mishra, B. Brahma, and A. K. Bhoi, “Depression Assessment in Youths Using an Enhanced Deep Learning Approach,” Studies in Computational Intelligence, vol. 1108, pp. 287–299, 2023, doi: 10.1007/978-3-031-38281-9_13.
Isam Eldin HA Magid et al., “Symptoms and risk factors for depression among college Freshmen students,” World Journal of Advanced Research and Reviews, vol. 14, no. 1, pp. 601–606, Apr. 2022, doi: 10.30574/WJARR.2022.14.1.0379.
O. Ya. Andriichuk and L. Y. Sakharuk, “ANALYSIS OF THE SOMATIC COMPONENT OF DEPRESSION IN THE DEVELOPMENT OF A PHYSICAL THERAPY PROGRAM,” Art of Medicine, pp. 12–17, Oct. 2022, doi: 10.21802/ARTM.2022.3.23.12.
M. Moran, “Academic Stress, Failure May Be Underrecognized Trigger for Suicidal Thoughts, Behaviors,” Psychiatr News, vol. 57, no. 8, Aug. 2022, doi: 10.1176/APPI.PN.2022.08.8.34.
J. Premalatha, S. Aswin, D. JaiHari, and K. Karamchand Subash, “Detecting Depression Using Quality-of-Life Attributes with Machine Learning Techniques,” Lecture Notes in Networks and Systems, vol. 728 LNNS, pp. 29–37, 2024, doi: 10.1007/978-981-99-3932-9_4.
K. Sujigarasharma, M. L. Shri, and K. Santhi, “Smart Detection and Proactive Prevention of Student Depression: A DeepLearning Approach,” TQCEBT 2024 - 2nd IEEE International Conference on Trends in Quantum Computing and Emerging Business Technologies 2024, 2024, doi: 10.1109/TQCEBT59414.2024.10545088.
M. Gupta, “A Comparative Study on Supervised Machine Learning Algorithm,” Int J Res Appl Sci Eng Technol, vol. 10, no. 1, pp. 1023–1028, Jan. 2022, doi: 10.22214/IJRASET.2022.39980.
Y. Zhou, Z. Zhang, Q. Li, G. Mao, and Z. Zhou, “Construction and validation of machine learning algorithm for predicting depression among home-quarantined individuals during the large-scale COVID-19 outbreak: based on Adaboost model,” BMC Psychology 2024 12:1, vol. 12, no. 1, pp. 230-, Apr. 2024, doi: 10.1186/S40359-024-01696-8.
H. Liu et al., “Depression Detection via Capsule Networks with Contrastive Learning,” Proceedings of the AAAI Conference on Artificial Intelligence, vol. 38, no. 20, pp. 22231–22239, Mar. 2024, doi: 10.1609/AAAI.V38I20.30228.
M. Ennab and H. McHeick, “Designing an Interpretability-Based Model to Explain the Artificial Intelligence Algorithms in Healthcare,” Diagnostics, vol. 12, no. 7, p. 1557, Jul. 2022, doi: 10.3390/DIAGNOSTICS12071557.
M. Balbino, R. Santana, M. Teodoro, M. Song, L. Zárate, and C. Nobre, “Predicting Depression in Children and Adolescents using the SHAP Approach,” pp. 514–521, Feb. 2022, doi: 10.5220/0010842500003123.
“Prediction of depression status in college students using a Naive Bayes classifier based machine learning model | Request PDF.” Accessed: Nov. 27, 2025. [Online]. Available: https://www.researchgate.net/publication/372683794_Prediction_of_depression_status_in_college_students_using_a_Naive_Bayes_classifier_based_machine_learning_model
E. O. Ogunseye, C. A. Adenusi, A. C. Nwanakwaugwu, S. A. Ajagbe, and S. O. Akinola, “Predictive Analysis of Mental Health Conditions Using AdaBoost Algorithm,” ParadigmPlus, vol. 3, no. 2, pp. 11–26, Aug. 2022, doi: 10.55969/PARADIGMPLUS.V3N2A2.
“Predictive Machine Learning Models for Assessing Lebanese University Students’ Depression, Anxiety, and Stress During COVID-19 - Christo El Morr, Manar Jammal, Imad Bou-Hamad, Sahar Hijazi, Dinah Ayna, Maya Romani, Reem Hoteit, 2024.” Accessed: Nov. 27, 2025. [Online]. Available: https://journals.sagepub.com/doi/10.1177/21501319241235588
D. Wang, Q. Li, Y. Gao, Y. Liu, and Y. Li, “Exploring the interpretability in speech-based adolescent depression detection by SHAP,” ACM International Conference Proceeding Series, pp. 562–567, Dec. 2023, doi: 10.1145/3638884.3638972.
D. Enkhbayar et al., “Explainable Artificial Intelligence Models for Predicting Depression Based on Polysomnographic Phenotypes,” Bioengineering, vol. 12, no. 2, Feb. 2025, doi: 10.3390/BIOENGINEERING12020186.
R. J. Kiran, J. Sanil, and S. Asharaf, “A Novel Approach for Model Interpretability and Domain Aware Fine-Tuning in AdaBoost,” Human-Centric Intelligent Systems, vol. 4, no. 4, pp. 610–632, Sep. 2024, doi: 10.1007/S44230-024-00082-2.
F. Ferren and F. I. Kurniadi, “Enhancing Bankruptcy Prediction with Feature Selection in AdaBoost Algorithm,” 10th International Conference on ICT for Smart Society, ICISS 2023 - Proceeding, 2023, doi: 10.1109/ICISS59129.2023.10291988.
Y. Lin, C. Li, and H. Li, “Machine learning-driven risk prediction and feature identification for major depressive disorder and its progression: an exploratory study based on five years of longitudinal data from the US national health survey,” J Affect Disord, vol. 381, pp. 573–583, Jul. 2025, doi: 10.1016/J.JAD.2025.04.056.
M. N. Fauzan, A. G. Putrada, N. Alamsyah, and S. F. Pane, “PCA-AdaBoost Method for a Low Bias and Low Dimension Toxic Comment Classification.,” ICACNIS 2022 - 2022 International Conference on Advanced Creative Networks and Intelligent Systems: Blockchain Technology, Intelligent Systems, and the Applications for Human Life, Proceeding, 2022, doi: 10.1109/ICACNIS57039.2022.10055017.
A. G. Putrada, M. Abdurohman, D. Perdana, and H. H. Nuha, “NoCASC: A Novel Optimized Cost-Complexity Pruning for AdaBoost Model Compression on Edge Computing-Based Smart Lighting,” COMNETSAT 2024 - IEEE International Conference on Communication, Networks and Satellite, pp. 729–734, 2024, doi: 10.1109/COMNETSAT63286.2024.10862047.
A. N. Iman, A. G. Putrada, S. Prabowo, and D. Perdana, “Peningkatan Kinerja AMG8833 sebagai Thermocam dengan Metode Regresi AdaBoost untuk Pelaksanaan Protokol COVID-19,” Jurnal Elektro dan Telekomunikasi Terapan, vol. 8, no. 1, p. 978, Aug. 2021, doi: 10.25124/JETT.V8I1.3894.
G. Grillo, T. Torda, C. Voena, A. Ciardiello, and S. Giagu, “Integrating ChatGPT-4: A Novel XAI Interface for Enhanced Clinician Understanding of MRI Image Segmentation Results,” Proc IEEE Symp Comput Based Med Syst, pp. 320–325, 2024, doi: 10.1109/CBMS61543.2024.00060.
“(PDF) Local Interpretable Model Agnostic Shap Explanations for machine learning models.” Accessed: Nov. 27, 2025. [Online]. Available: https://www.researchgate.net/publication/364522755_Local_Interpretable_Model_Agnostic_Shap_Explanations_for_machine_learning_models
