Liver Disease Detection Due to Excessive Alcoholism Using Data Mining Techniques

Alcohol is consumed in excess by millions of people across the world. Alcohol consumption is directly linked to life threatening liver diseases such as cirrhosis which may ultimately lead to death. Early detection of liver disease caused by over consumption of alcohol would help in saving lives of many people. By detecting liver disease in its early stage, it can be diagnosed in time and may lead to full recovery in some patients. This paper proposes detection as well as to predict the presence of liver disease using data mining algorithms. We will make a decision tree for the dataset and then the rules will be generated. After determining the rules, we will use different data mining algorithms to train and test the dataset to detect the liver disease. The data was collected from UCI repository and our training dataset was developed. It consists of 7 different attributes having 345 instances. In the dataset, different categories of blood tests are taken into considerations which are directly linked to liver diseases that may arise due to excessive alcohol consumption along with frequency of alcohol consumption. Based on the type of liver disease detected, prognosis may be suggested.