A Supervised Learning-Based Machine Learning Method for Detecting Depression and Anxiety

Abstract

Anxiety and depression are two of the main factors contributing to significant disability in emerging nations. India ranked top in the World Health Organization's (WHO) South East Region report on anxiety disorders, with women suffering from the condition twice as badly as men. Since treating these problems later on would be more expensive and ineffective than intervening early, we have developed a model that diagnoses the various stages of these mental disorders using machine learning algorithms in conjunction with a regular psychological assessment. We've identified five distinct AI algorithms—Convolutional Neural Networks, Support Vector Machines, Linear Discriminant Analysis, K Nearest Neighbor Classifiers, and Linear Regression— that have proven effective in our proposed model when applied to both anxiety and depression datasets. In assessing diverse measurement criteria like accuracy, recall, and precision, our model employing the CNN algorithm demonstrates superior performance. Specifically, it achieves 96% accuracy for anxiety and 96.8% for depression, surpassing other algorithms in these aspects. Furthermore, our data reveals that 15.6% of urban women aged 18 to 35 experience chronic depression, and 7.4% experience severe anxiety.