Early Prediction of Chronic Kidney Disease using Machine Learning Techniques

The proposed study uses a systematic strategy that includes multiple important elements to classify chronic kidney disease (CKD). First, the dataset with all the necessary characteristics loaded (age, sex, antivirals, steroids, weariness, malaise, anorexia). Data cleaning techniques are then applied to guarantee the dataset's dependability and integrity for analysis. Then, feature selection techniques are used to determine which attributes have the greatest bearing on the classification of CKD. After that, the dataset is subjected to a variety of classification techniques, such as Naive Bayes, Decision Tree, Kstar, Logistic Regression, and Support Vector Machine (SVM), to assess how well they perform in correctly categorizing cases of chronic kidney disease. The findings show that all methods have the same level of classification accuracy, with Logistic Regression performing the best at 86.452%.