Classification of Depression Disorder Using Fuzzy C Means

https://doi.org/10.46610/JOSME.2023.v09i01.003

Abstract

Depression is a diverse syndrome in which some underlying presentations may share a common phenomenology but have different etiologies. Despite considerable work on the etiology of depression including neurobiological, genetic, and psychological studies, no reliable classificatory system has emerged that links either to the underlying etiology or has proven strongly predictive of response to treatment. Reactive and endogenous depression, melancholia, typical depression, depression with a seasonal pattern/seasonal affective illness, and dysthymia are only a few of the classification systems/subgroupings that have been employed. These have been based on a variety of factors, including the type, quantity, intensity, pattern, and duration of the symptoms, as well as in some cases, the presumptive cause.

We adopted the Beck Depression Inventory (BDI)-II as the instrument, and outpatients of a psychiatric clinic were recruited as samples, as well as undergraduates as a non-clinical sample, to achieve the goal of studying the classification of depression disorders using fuzzy theory. The elements in the BDI were presented, and percentages were given to each one. We have the option to pick from many membership levels, and the total percentages for the selected category might reach 100%. To categorize the severity of depression, we use probability-based categorization Wald's and k-means together with fuzzy C-Means. We categorize the FCM, Wald's, and k-means percentages using clustering analysis. Finally, we conclude that FCM performed better than the other two probabilities-based techniques.