Arif, HossainMobin, Md. IftekharulShawan, Naveed RahmanMehrab, Syed Samiul AlamAhmed, FardeenHasmi, Mohammad Sharatul2019-06-252019-06-2520192019ID 13201027ID 16101175ID 15101073ID 15101133http://hdl.handle.net/10361/12255Cataloged from PDF version of thesis.Includes bibliographical references (page 40).This thesis is submitted in partial fulfillment of the requirements for the degree of Bachelor of Science in Computer Science and Engineering, 2019.Chronic kidney disease (CKD) is the gradual loss of kidney function over a duration of months or years. One in ten people are affected by it at some stage. Some ethnicities such as African Americans and South Asians are predisposed to having the disease. Globally the number of people affected has been growing through the years, with 752.7 million having the disease in 2016 The disease has no cure, so early detection is key to better manage the disease and control other risk factors such as diabetes and blood pressure. Although CKD has no early symptoms and requires medical tests on blood and/or urine samples, medical tests conducted for other diseases hold clues to whether someone has CKD . The datasets that are available have a multitude of features and are also incomplete and imbalanced. We want to overcome this problems through feature engineering to reduce the number of features. A comparative study of various classifiers needs to be done to find those that hold promise and are robust enough to handle currently available datasets, which are both incomplete and unbalanced. Our study is to bring down the number of attributes/features using recursive feature elimination method and use Ensemble classifier to predict the existence of CKD from the reduced features.41 pagesenBrac University theses are protected by copyright. They may be viewed from this source for any purpose, but reproduction or distribution in any format is prohibited without written permission.Ensemble learningImbalanced datasetChronic kidney diseaseMachine learningMachine learningThesis