Estimation of Factors Related to Premenstrual Syndrome in Female Students by Using Artificial Neural Network Model

Authors : Yeşim AKSOY DERYA, Cemil ÇOLAK, Burcu KAYHAN TETİK, Ayşe Nur YILMAZ, Zeliha ÖZŞAHİN

Pages : 88-93

View : 12 | Download : 8

Publication Date : 2020-12-31

Article Type : Research Paper

Abstract :Objective: Premenstrual syndrome is a disorder with psychological and physical symptoms in almost every cycle, and it concerns almost 90% of women of reproductive age. In recent years, studies investigating the relationship between premenstrual syndrome and vitamin D, trace elements, and lymphocyte/platelet ratios have been conducted. In this study, the factors associated with PMS in female students were tried to be determined by using an artificial neural network insert ignore into journalissuearticles values(ANN); model. Materials and Methods: This study was conducted on female students at Inonu University Faculty of Medicine and Health Sciences, between 01 May and 30 June 2019. Demographic characteristics and menstruation histories of 860 female students were collected and recorded for the study. A multi-layer perceptron artificial neural network model was used to determine the factors associated with premenstrual syndrome. The performance of the model is determined by the accuracy rate and the area under the process characteristic curve. Results: Correct classification rates of the created multi-layer perceptual neural network model for premenstrual syndrome were calculated to be 63.2% in the training data set and 63.0% in the test data set. Considering the importance values of the variables; it was found that the duration of active internet use insert ignore into journalissuearticles values(phone, tablet, computer); was the most influential factor on premenstrual syndrome and the economic status of a student was the least ,influential factor. Conclusion: According to the findings of the designed artificial neural network model, the three most important factors related to premenstrual syndrome were determined to be the duration of active Internet use, present age and age of menarche. Given the high prevalence of PMS, the uncertainty of etiology, and its potential to affect a woman`s lifestyle; the use of artificial intelligence models with larger sample size and including different factors is recommended.
Keywords : Artificial neural network model, Associated factors, Female student, Premenstrual syndrome