Analytical Comparison for the Diagnosis of Chronic Kidney Disease Applying Intelligent ML-based Systems

Diseases like Chronic Kidney Disease (CKD) develop gradually over time and then stick around for a very long time. At its most advanced stage, it is fatal, and the only treatment option is a transplanted kidney or ongoing dialysis, which is an artificial filtering mechanism. Even though there have been improvements in surgical care and treatment, CKD remains a Serious Threat to Health. As CKD has become more common in recent years, scientists throughout the world have been increasingly invested in finding effective ways to diagnose, treat, and prevent the disease. By studying the features that cause the problem, it is possible to improve performance. Healthcare providers can learn more about a patient’s condition before a visit to the clinic by analyzing the patient’s medical records. Although an early diagnosis of CKD is crucial for providing the curative or at least manageable treatments it requires, the effectiveness of current systems remains to be improved. A clever categorization and prediction model is proposed in this paper. Classification techniques such as Artificial Neural Networks (ANNs), XGBoost, and K-NN are used to provide predictions about kidney illness. Based on the results of the evaluation, the suggested model outperforms the state-of-the-art methods for predicting CKD by 98.33 percent. Analyzing these outcomes demonstrates that cutting-edge ML learning approaches are useful for clinical decision-making and can aid in the early prediction of CKD and its related phases, which can help to slow the worsening of kidney damage. Algorithms for the dataset obtained from the default UCI repository have been compared using the alternative metric. This study also emphasizes the significance of statistical analysis and domain understanding of the features when extrapolating from clinical data to predict CKD.