Mental Health Prediction in Students using Data Mining Techniques

Barnali Sahu1, *, Jahnavi Kedia1, Vaishnavee Ranjan1, Biranchi Prasad Mahaptra1, Satchidananda Dehuri2
1 Department of Computer Science and Engineering, Siksha ‘O’ Anusandhan Deemed to be University, Bhubaneswar, Odisha, India
2 Department of Information and Communication Technology, Fakir Mohan University (State Government University), Balasore, India

Article Metrics

CrossRef Citations:
Total Statistics:

Full-Text HTML Views: 852
Abstract HTML Views: 541
PDF Downloads: 260
ePub Downloads: 177
Total Views/Downloads: 1830
Unique Statistics:

Full-Text HTML Views: 563
Abstract HTML Views: 201
PDF Downloads: 204
ePub Downloads: 139
Total Views/Downloads: 1107

Creative Commons License
© 2023 Sahu et al.

open-access license: This is an open access article distributed under the terms of the Creative Commons Attribution 4.0 International Public License (CC-BY 4.0), a copy of which is available at: This license permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.

* Address correspondence to this author at the Department of Computer Science and Engineering, Siksha ‘O’ Anusandhan Deemed to be University, Bhubaneswar, Odisha, India; E-mail:



Mental health issues are common among university students. Depression poses a significant obstacle to long-term learning and the viability of the educational system as a whole. Many students enroll in colleges in other states, leaving their friends and family behind. Some students struggle to adjust to competition in a foreign setting and the pressures of college. With numerous technical and non-technical universities, Odisha is currently rising as the center of education. There has been much research that has examined mental health issues in India, but few of them specifically target university students in Odisha. Our study aimed to predict the prevalence, factors, and effects of mental health problems (depression, social connectedness, and anxiety disorder) on college students in Odisha.


An online survey was conducted that was circulated to several student classes at 3 colleges of an university, and it yielded 109 results. The survey included socio-demographic information along with the General Anxiety Disorder Questionnaire (GAD-7), the Revised Social Connectedness Scale (SCS), and a nine-item Patient Health Questionnaire (PHQ-9) scale. Correlation analysis has been applied to identify the correlation among attributes and regression analysis was applied for the prediction of the mental health status of the students based on the given attribute.


The prevalence rate of depression among students was determined to be 61.90%. Years of college and physical health showed a significant correlation with depression. Students in the early years of college have shown a greater depression rate. It was observed that anxiety and depression were positively correlated and social connectedness and depression were negatively correlated. We also found academic performance and depression to be correlated with each other. The hyperparameter-tuned logistic regression model provided better result in comparison to the other existing models in the literature.


The findings hint at the high prevalence of depression in students and its association with anxiety disorders, social connectedness, and academic performance. This study emphasises how crucial it is for decision-makers to develop preventative measures and policies for a sustainable educational system.

Keywords: Mental health, Depression, Social connectedness, Data mining, Correlation, Prediction models.