Performance Analysis of Classification Algorithms on Liver Disease Detection

Liver is the prime limb in the body and executes a number of functions related to producing necessary proteins, carbohydrates metabolism, energy storage, and detoxification of waste. Diseases that may affect the liver include hepatitis, fatty liver, bleeding, fatigue, jaundice. Liver disease affects millions of the populace and diagnosis at an initial phase is indispensable for better handling. So there is a solid requirement for an efficient, accurate, and practical framework to predict the consequence of such infection. It will be valuable for taking precautions and appropriate treatment. To reduce this concern, we proposed a system that is capable enough to detect liver malady using machine learning. The major intention of this work is to use classification algorithms to recognize the liver patients from healthy individuals with the help of automated processes to overcome human fault and decrease the false alarm rate that gives the progression which is more accurate, effortless, and unproblematic. This project also aims to compare the classification algorithms that depend on their performance features. To gain results with less processing instance and a high precision rate. In this work, four classification algorithms that are LR, RF, KNN, and DT have been applied for analyzing their performance which is based on the liver patient data. The dataset contains 583 liver images and KNN was the model that generated a great accuracy with 72.04% accuracy.