Main Article Content
One of the most vital causes of death worldwide is liver disease. We, humans, have come a long way in the medical field and scientific advancements to treat diseases and it's evident that when these liver diseases are detected early, they can be treated easily. In order to be able to accurately predict if there’s a chance of the liver disease it is imperative to identify the features/symptoms which play a significant role in causing the Liver Disease. In order to improve the performance of the prediction models, it is important to choose the right combination of significant features.
A new system is proposed that identifies the significant features and then predicts whether or not a person may suffer or is suffering from Liver Disease using the identified features. Our system ought to be used as a supplementary tool in diagnosis. Data is essential and we will be using the dataset available on the UIC repository. We will be using genetic algorithms to identify the significant features and then use those features to train different classification models like k-Nearest Neighbors, k-means, Random Forest, Support Vector Machines, Naïve Bayes, Logistic Regression, etcetera which will predict if there’s a chance of Liver Disease for a person’s data. We will also be using neural networks with back propagation to perform binary classification.
Ideally our proposed model identified the significant features and finds the best model which predicts the Liver Disease with more accuracy or another statistical measure.