Classification of Student Mental Health Analysis using Logistic Regression and other classification techniques through Machine Learning Methods

• Published in: 2024 3rd International Conference for Innovation in Technology (INOCON) pp. 1 - 5
• Main Authors: Pritam, Nishant, Gill, Kanwarpartap Singh, Kumar, Mukesh, Rawat, Ruchira, Banerjee, Deepak
• Format: Conference Proceeding
• Language: English
• Published: IEEE 01.03.2024
• Subjects: Analytical models; anxiety; Artificial intelligence; classifier; Data Analytics; Decision Tree; depression; Linear regression; Logistic regression; Machine learning; Machine learning algorithms; Mental health; modelling; Random Forest; Statistical analysis; Students; Support vector machines; SVM
• DOI: 10.1109/INOCON60754.2024.10512216

The examination of student mental health is a significant area of scholarly investigation that seeks to comprehend and address the psychological and emotional well-being of students within educational settings. The process involves the evaluation, anticipation, and provision of assistance for students' psychological well-being using various approaches, including as surveys, machine learning algorithms, and clinical assessments. The use of machine learning algorithms for the analysis of student mental health is a complex and consequential endeavour. This may provide insight on the identification of students who may be at risk and the appropriate timing for providing support. In this context, the use of machine learning methods, such as logistic regression and other classification algorithms, might potentially be advantageous. This research aims to identify some noteworthy concerns pertaining to the overall mental well-being of students. This paper explores the philosophical difficulties underlying these challenges and examines the answers provided by modern statistics and visualisation techniques. This study offers a variety of robust models, including as Random Forest, Decision Tree, SVM, and Logistic Regression, that possess several advantages and are well-suited for the analysis of categorical data. The research use a psycholinguistic data set to conduct a comprehensive comparison of different statistical methodologies. Upon conducting an evaluation of the classification models, including Linear Regression, Support Vector Machine (SVM), Decision Tree, and Random Forest, it was seen that the Linear Regression Classification Technique exhibited the highest level of accuracy. Specifically, the Linear Regression model achieved a 65 percent accuracy rate across diverse optimisation parameters.