Diagnosis of Liver Disease Induced by Hepatitis Virus Using Machine Learning Methods

Liver, the largest vital organ in the human body, helps in digestion, energy storage, and toxic removal processes. Hepatitis A, B, C, fatty liver, cirrhosis, and liver cancer all cause swelling of the belly and legs, changes in the color of urine, and jaundice. Early identification is necessary to save the lives of those infected with the Hepatitis virus. Although clinical tests such as Abdominal Ultrasound, Chloride Blood Test, Liver Biopsy, Liver Function Tests (LFT), Magnetic Resonance Cholangio Pancreatography (MRCP), Total Protein and Albumin/Globulin (A/G) Ratio, and others are used in the diagnosis of Hepatitis virus, but high-level expertise of the doctors are also required. With continual advancements in Artificial Intelligence (AI) and Machine Learning algorithms, medically intelligent systems for early detection of fatal liver disease can be developed for prediction of deadly liver disease in the early stage. The goal of this paper is to examine the accuracies of several Machine Learning algorithms such as Linear Regression, Decision Tree, Random Forest, KNN, and SVM. For this work, author used the data set accessible at the UCI database and proved that the Support Vector Machine perform better than all other methods with.91.84% accuracy.