Machine Learning Algorithms for Predicting Factitious Disorder Using the Learning Vector Quantization Method

Abstract

Factitious disorder is classified as a mental problem because it is related to severe emotional disorders. A person who has factitious disorder intentionally produces symptoms of the disease for the purpose of receiving care and attention in a medical setting. People with Factitious disorder act with the aim of attracting the sympathy and attention of others. Diagnosing factitious disorder is very difficult. The reason is, the sufferer looks fine. The doctor must eliminate any physical and mental illness before confirming the diagnosis of Factitious disorder. Along with the development of machine learning technology. Incorporation of patient data and the use of machine learning technology can help detect the disease. The purpose of this study was to build a system to predict the likelihood of a person being exposed to factual or unrelated disorders with the inputs that the patient entered. The method for diagnosing factitious disorder uses the Learning Vector Quantization method whether a person is a sufferer or not. The data was obtained from the questionnaire using 14 parameters and managed to get data as much as 30 training data. This research resulted in a maximum epoch value of 1000, a learning rate value of 0.1, a learning rate multiplier of 0.1, a minimum learning rate of 0.0001 and a training data of 5. The accuracy result obtained is 70%.