chronic kidney disease (CKD), Data mining, Bagging, Random subspace Naive Bayes, KNN, Decision tree, Accuracy, ROC, Kappa
Abstract
chronic kidney disease (CKD) is a global health issue that causes a high incidence of morbidity and death, as well as the onset of additional illnesses. Because there are no clear symptoms in the early stages of CKD, people frequently miss it. Early identification of CKD allows patients to obtain prompt therapy to slow the disease's development. Due of their rapid and precise identification capabilities, machine learning models can successfully assist doctors in achieving this aim. We present a machine learning framework for diagnosing CKD in this paper. The CKD data set was taken from the machine learning repository at the University of California, Irvine (UCI), which has a substantial number of missing values. To fill up the missing values, KNN imputation was employed, which picks many full samples with the most comparable measurements and processes the missing data for each incomplete sample. For a variety of reasons, patients may miss some measurements, resulting in missing statistics in real-world medical settings. Six machine learning approaches (logistic regression, random forest, support vector machine, k-nearest neighbour, naive Bayes classifier, and feed forward neural network) were employed to create models once the missing data set was filled in appropriately. Random forest outperformed the other machine learning models, with a diagnostic accuracy of 99.75 percent. After examining the misjudgments produced by the current models, we proposed an integrated model that incorporates logistic regression, random forest, and perceptron, which could achieve an average accuracy of 99.83 percent after ten simulations. As a result, we hypothesized that this technology may be used to diagnose diseases using more complex clinical data.
Article Details
Unique Paper ID: 154715
Publication Volume & Issue: Volume 8, Issue 12
Page(s): 158 - 168
Article Preview & Download
Share This Article
Conference Alert
NCSST-2023
AICTE Sponsored National Conference on Smart Systems and Technologies
Last Date: 25th November 2023
SWEC- Management
LATEST INNOVATION’S AND FUTURE TRENDS IN MANAGEMENT