COVID-19 impact on students' mental health: explainable AI and classifiers

Abstract

COVID-19, lockdown, and isolation have included an enormous impact on the students around the world like others. As isolation strategy with quarantine is useful to prevent transmission, students remaining at home gained nothing but illness perception, anxiety, and depression in spite of sharpening their knowledge and reflecting the thoughts. The detachment from routine life has affected the pillars of the mental health balance and isolated and suffocating lives have created toxic feelings in lives. Therefore the purpose of our paper is to predict the mental health of students in such situations. To accomplish our work, we have collected the students' mental health survey dataset from the Kaggle website later trained the data with suitable classifiers to predict mental health. In this paper, we demonstrated five different classifiers models to predict optimal accuracy, including two different Explainable AI (XAI) techniques (LIME, SHAP) as it enhances the trust in an AI system.