A Comparative Study of Classification Methods for Predicting Chronic Kidney Disease
Now a days predicting diseases in healthcare has become one of the important task. Machine learning takes a major role for prediction and classification purposes in medical field. Chronic Kidney Disease (CKD) is resulted as one of the most basic health issue as a result of its developing pervasiveness. In India every year approximately 1 million people affected by CKD. CKD is a disease which is caused by harm to the two kidneys. Chronic Kidney Disease joins the state where the kidneys neglect to work and decrease the possibility to keep an individual suffering from disease. Early recognition and appropriate medications can avoid or decrease the movement of this chronic kidney ailment to conclusive stage, where as kidney transplantation or dialysis is the simple way to survive life. Data mining is one of the present key process used in performing analytic outcomes. Data mining techniques are used which helps in discovering useful data from huge datasets which are available from human health industry. The paper aims at early predicting the presence of CKD by utilizing machine learning strategies. In order to evaluate our approach we consider CKD dataset of 400 patient individuals contains of 25 attributes. By considering features selection on CKD dataset we perform KNN, SVM, Random Forest algorithms. Based on accuracy we compared different machine learning algorithms that will help people in predicting the presence of CKD or not.