Abstract

Depression remains a significant global challenge, being one of the most prevalent and costly mental disorders that adversely affects the quality of life, as supported by numerous studies. Enhancing our understanding of the factors linked to quality of life is crucial to optimize long-term outcomes and diminish disability in individuals with depression. The primary focus of this study centers around recognizing depression based on lifestyle and livelihood factors. It is important to note that depression can impact individuals of all ages, genders, and backgrounds, often stemming from a combination of genetic, biological, environmental, and psychological factors. Additionally, major life events, chronic stress, trauma, or a family history of depression can contribute to its development.

In the healthcare sector, machine learning methodologies are utilized to process and interpret diverse data types, with the goal of gaining deeper understanding of the relationship between quality-of-life indicators and depression. Different classification algorithms such as Random Forest, Decision Tree, Naive Bayes, Support Vector Machine, and PPMCSVM have been leveraged for this analytical purpose.

Hence, the main objective of this proposed work is to improve depression prediction by utilizing an ensemble technique that identifies the factors associated with quality of life among depressed patients. To achieve this, the study adopts KNN (K-Nearest Neighbour) and Voting Classifier algorithms. The Voting Classifier facilitates the identification of the underlying causes of depression in an individual. The findings from this research demonstrate that the suggested model can accurately predict the causes of depression, paving the way for more effective intervention and treatment strategies.

Keywords: Depression, mental disorder, Healthcare System, PPMCSVM, prediction accuracy, ensemble technique, KNN (K-Nearest Neighbour), Voting Classifier, underlying causes, intervention, treatment strategies.