This document discusses using data mining classifiers and attribute reduction techniques to predict chronic kidney disease (CKD) more accurately and efficiently. It first provides background on CKD and the need for early detection. It then discusses data mining, classification algorithms, attribute selection filters and wrappers. The document analyzes several studies that predicted CKD using techniques like decision trees, SVM and Naive Bayes. It describes the dataset used from the UCI repository and evaluation metrics. The results section compares J48, Decision Tree and IBK classifiers with and without attribute reduction using CfsSubsetEval, ClassifierSubsetEval and WrapperSubsetEval. Attribute reduction improved accuracy, especially for IBK which achieved 100% accuracy with 72% fewer attributes.