ANN Model for Liver Disorder Detection

Liver plays a significant role and hence considered to be the major organ in human body. Diagnosis of any disorder in liver at initial stage is vital for its recovery. Liver diseases occur because of high rate of alcohol consumption, consumption of contaminated water or food and being exposed to toxic gases, etc. Early prediction of the disease can save a life. Evaluation of the classification algorithms: Logistic Regression, Support Vector Machine, Naive Bayes, and Artificial Neural Network are done in this paper by carrying out the experimental approach in Anaconda by using “sklearn” and “keras” libraries. The dataset has been acquired from UCI Machine Learning repository with 10 major attributes. The aim is to find the model which most accurately predicts any disorder in the liver. Through our experiment, Artificial Neural Network has the best fit with accuracy of 80.70%, making it better than other techniques.