Klasifikasi Kesehatan Mental Mahasiswa Menggunakan Light Gradient Boosting Machine Dan Analisa Fitur Menggunakan SHAP

Abstract

The mental health of college students is an important issue as many do not receive treatment despite needing it. According to the Association of University and College Counseling Center Directors 95% of college students experience an increase in psychopathology. This study uses the Light Gradient Boosting Machine algorithm to classify the mental health of college students based on a dataset that has a total of 61.794 rows and 16 columns. Light Gradient Boosting Machine is an implementation of Gradient Boosting Decision Tree which has two strategies namely gradient-base one-side sampling (GOSS) and leaf-wise growth. The accuracy results obtained using LightGBM reached 67% where the data used had been balanced using the class_weight parameter and the ADASYN technique. In addition, the research was analyzed to find the most contributing features using the SHAP (SHapley Additive exPlanations) method with the results obtained there are 6 features that have the highest contribution value including Country, treatment, mental_health_interview, family_history, Gender, dan self_employed.