International Research Journal of Engineering and Technology (IRJET)
e-ISSN: 2395-0056
Volume: 10 Issue: 08 | Aug 2023
p-ISSN: 2395-0072
www.irjet.net
Mental Illness Prediction using Machine Learning Algorithms Falguni Wani1, Ved Deore2, Shivam Gorane3, Santosh Chobe4 1,2,3Student, Dept. of Computer Engineering, Pimpri Chinchwad College of Engineering and Research, Ravet, Pune,
Maharashtra, India
4Professor, Dept. of Computer Engineering, Pimpri Chinchwad College of Engineering and Research, Ravet, Pune,
Maharashtra, India ---------------------------------------------------------------------***--------------------------------------------------------------------Mental health is an integral and essential Abstract - Depression is one of the most concerned issues in
component of health. The WHO constitution states: "Health is a state of complete physical, mental and social well-being and not merely the absence of disease or infirmity."[2]. The leading symptoms of depression from a medical point of view are lack of concentration, loss of memory, loss of interest in recreational activities, an inability to make decisions, overeating and weight gain, weight loss, low appetite and irritation, etc. These symptoms have a significant effect on crucial areas of an individual’s life.
the society and it is not limited to certain age of a person. Depression management is an approach for analyzing and working on these concerns and lead to quality of life. The idea behind this work is to analyze depression, anxiety and stress based on some psychological test like Depression Anxiety Stress Scale-21(DASS 21). Machine learning is an emerging field in computer science and has ability to predict outcome based on certain situations or inputs. Machine learning algorithms are used to predict depression, anxiety and stress levels by using standard psychological scale. Training and testing datasets are used to train and test the developed machine learning model. Various machine learning algorithms like Support Vector Machine, Random Forest, Naïve Bayes, etc. are implemented and compared in order to evaluate the best among all. The accuracy of the best algorithm is boosted using the boosting technique of ensemble learning method and a user interface is used for self-evaluation. From the classification algorithms used SVM has surpassed the other machine learning algorithms and then it is boosted using AdaBoost giving highest accuracy for prediction.
The symptoms of anxiety are irritability, insomnia, nervousness, sweating, fatigue, panic, increased heart rate and a sense that something is about to happen, difficulty in concentrating and rapid breathing. The common symptoms of stress are low energy levels, feeling upset or agitated, chronic headaches, impotence to relax, recurring overreaction and persistent colds or infections. Thus, anxiety, stress and depression have many common symptoms including fatigue, chest pain, insomnia, inability to concentrate and increased heart rate all of which makes classification tough for machines.
Key Words: Depression, Anxiety, Stress, Classification, Depression Anxiety Stress Scale-21(DASS-21), Supervised Learning, AdaBoost, Support Vector Machine
This paper is structured as follows: Section 2 explores related studies on anxiety, depression and stress along with the methods and techniques that were adopted. Section 3 describes the dataset used in the research herein, while Section 4 discusses the various classification algorithms. Section 5 studies the research gap found. Section 6 includes experimental setup used to perform this study, while section 7 describes the proposed system. Section 8 compares the results of machine learning algorithms. Finally, section 9 is the conclusion, which summarizes the study in its entirety.
1. INTRODUCTION Healthcare is among the serious issues in front of the entire world regardless of any circumstances. As a ruling interest globally, besuited, well organized, effective and robust wellness systems are built to improve and conserve the quality standards of life. Anxiety, depression, stress, irritation and disappointment have become so normal that individuals now imagine them to be part of personal and professional life.
2. LITERATURE REVIEW
The World Health Organization (WHO) has estimated that 3.8% of the population experience depression, including 5% of adults (4% among men and 6% among women), and 5.7% of adults older than 60 years. Approximately 280 million people in the world have depression [1]. Differentiating between anxiety and depression is complicated for machines; therefore, a suitable machine learning algorithm is necessary for an applicable recognition.
© 2023, IRJET
|
Impact Factor value: 8.226
The literature survey shows the study of various machine learning algorithms to predict depression, anxiety and stress. In [3], Anu Priya, et. al. proposed machine learning model for predicting different levels of depression, anxiety and stress. They applied different machine learning algorithms like Support Vector Machine (SVM), Decision Tree (DT), Naïve Bayes (NB), Random Forest (RF) and KNearest Neighbors (KNN). They also calculated different
|
ISO 9001:2008 Certified Journal
|
Page 191