As a result of the rising incidence of the condition, chronic kidney disease (CKD) has emerged as one of the pressing issues facing modern medicine. The capability of a machine vision system to diagnose chronic renal disease based on a small number of patient characteristics is the focus of this investigation. A large number of statistical tests, such as ANOVA tests, Pearson's correlation test results, and Cramer's V tests, have indeed been carried out in order to get rid of duplicate features. All of the different machine learning algorithms—including regression models, random forest, support vector machine, and gradient boosting—were trained and evaluated through a 10-fold cross-validation method. In order to attain an efficiency of 99.53, we make use of the F1 metric that is associated with the gradient augmenting classifier. In addition, it was discovered that haemoglobin has a significant role in the differentiation of CKD in the randomized forest as well as the xgboost models. Our findings represent the most impressive, despite the fact that our study had fewer distinguishing features than those of earlier research. Because of this, chronic kidney disease can be diagnosed with just three easy tests that cost a total of $26.65 just.
Key words: Machine , Learning , Approach , Kidney , Disease , Prediction
|