Prognosis of Liver Disease: Using Machine Learning Algorithms

Abstract

The process of identifying patterns in huge datasets comprising methods such as machine learning, statistics, and database system can beconsidered data mining. It is a multidisciplinary field in computer science and it excerpts knowledge from the massive data set and converts into comprehensible format. The Medical environment is rich in information but weak in knowledge. Medical systems contain wealth of data which require a dominant analysis tool for determining concealed association and drift in data. The health care condition that comprehends to liver disorder is termed as Liver disease. Liver disorder leads to abrupt health status that precisely governs the working of liver and intern affecting other organs in the body. Data mining classification techniques like Decision Tree, Linear Discriminant, SVM Fine Gaussian and Logistic Regression algorithms are applied. Laboratory parameters of the patients are used as the dataset. Data contains features that can establish a rigorous model using Classification technique. MATLAB2016 is used in this paperfor implementing classification algorithm on the dataset. Linear Discriminant algorithm showed the highest prediction accuracy 95.8% and ROC is 0.93.

Ähnliche Arbeiten