Abstract: The abstract introduces the pressing issue of Chronic Kidney Disease (CKD) and underscores the importance of early identification to mitigate its progression and enhance patient outcomes. It highlights the increasing utilization of machine learning (ML) algorithms for CKD prediction but identifies a need for more accurate and efficient models. The paper aims to fill this gap by conducting a thorough literature review on CKD prediction using ML techniques, analyzing features, datasets, algorithms, and evaluation metrics utilized in existing studies. Additionally, it proposes a novel approach that combines different feature selection and ML techniques to improve prediction accuracy. The findings demonstrate the potential of ML algorithms such as support vector machines, random forests, and neural networks to achieve high accuracy in CKD prediction, with the proposed approach enhancing accuracy by up to 5%. The implications of this study suggest the development of more effective CKD prediction models that could positively impact clinical practice and patient outcomes.
Keywords:Chronic Kidney Disease (CKD), Machine Learning (ML), Prediction, Feature Selection, Datasets, AlgorithmsEvaluation Metrics, Support Vector Machines (SVM), Random Forests, Neural Networks, Accuracy Improvement, Clinical Practice, Patient Outcomes, Healthcare Management, Early Identification.
|
DOI:
10.17148/IJARCCE.2024.13378
