ESTIMATION OF DEPRESSION DISEASE BY NEURAL FUZZY INFERENCE METHODD

Authors : Zekeriya AKÇAY, Remzi GÜRFİDAN

Pages : 49-58

View : 24 | Download : 14

Publication Date : 2019-12-16

Article Type : Research Paper

Abstract :In this study, an ANFIS model that analyzes the data set of depression disease was established and the degree of disease was estimated in this study which was performed for the detection of depression disease by neural fuzzy logic inference method insert ignore into journalissuearticles values(ANFIS);. The data set was prepared according to Beck Depression Test results. The Neuro Fuzzy Designer included in Matlab R2016b was used to process the data set using ANFIS method and generate estimation values. In ANFIS, sugeno method was used and Trimf was selected as the activation function. The learning method is the backpropa method known as the back propagation method. At the end of 50 trainings, the training error value was determined as 0.018197. The training error value resulting from processing a total of 200 disease records is in good condition. The actual values and the estimated values produced by the system were analyzed with SPSS Statistics software and standard deviation and error values were determined. The system is aimed to classify the degree of disease correctly according to the symptoms.
Keywords : ANFIS, Fuzzy Logic, Artificial Neural Network, Depression, Disease