Liver Disease Prediction using Semi Supervised based Machine Learning Algorithm

Abstract

Recently, liver disease is emerging as one of the most common diseases with high fatality rate. The number of liver affected patients are steadily increasing due to various factors such as excessive alcohol consumption, inhalation of hazardous fumes, eating tainted food, pickles, and narcotics. Liver disease can also lead to a variety of serious illnesses, including liver cancer. To enhance the process of liver disease classification, a semi supervised machine learning algorithm has been presented in this research work. The use of liver patient datasets in the development of classification algorithms to predict liver disease is being investigated. The proposed study employs hybrid SVM and K-Means algorithm-based model to examine the entire patients’ liver disease. Chronic liver disease is defined as a liver ailment that exist for at least six months. As a result, the considered population will be classifiers as liver disease affected and non-affected people. The SVM classifiers are used to determine the number of liver disease affected individuals, and then the hybrid k means clustering method is used to determine how much of the liver has been affected. As a consequence, the output from the proposed hybrid K-Means clustering model reveals high prediction accuracy.