An Ensemble Approach to Chronic Kidney Disease’s Diagnosis

Abstract

Abstract: The most essential priority right now is healthcare which incorporates the identification, treatment, prevention, and management of illness, injury, or sickness. Chronic diseases, the most harmful type of diseases, are more prevalent in senior individuals and are often treatable but incur a significant financial cost, adding to the challenges the patient and the patient's family already confront. These have great impacts on the kidneys, damaging the waste-filtering mechanism of the body. As a result, technologies like artificial intelligence (machine learning-ML as well as deep learning-DL) are now being used to forecast and enhance the health of human systems in an efficient, inexpensive and reliable way. In this study, a stacking model with SVC, Adaboost and Random forest is being proposed which is trained on a dataset (n=400) collected in india over a time span of two months which includes 25 features (including red blood cell (RBC) count, white blood cell (WBC) count, etc). The data went through Exploratory Data Analysis (EDA), followed by feature extraction using Adaboost. This data was then used for model training using different classifiers, including the proposed model. The stacking model gave best accuracy (100%), precision (100%) and recall (100%) in comparison to SVC (Support Vector Classifier), Random forest and Adaboost models used individually.