CHRONIC KIDNEY DISEASE PREDICTION MODEL USING BAYESIAN OPTIMIZATION AND XGBOOST MACHINE LEARNING ALGORITHM
Keywords:
Bayesian Optimization, Chronic Kidney Disease, Hyperparameter Tuning, Machine Learning, Predictive Modeling, XGBoostAbstract
Chronic Kidney Disease or CKD is a global concern that continues to flourish and affect the wellbeing of people and systems across the world. It is defined by the gradual loss of kidney functionality that leads to important issues like cardiovascular disease, renal failure, and increased death rates. Previous researchers has concentrated towards the development of machine learning algorithms Random Forest, K Nearest Neighbour algorithm, Decision Tree and Deep Neural Network for CKD prediction, but higher prediction accuracy and model interpretability has not been achieved. Although some researchers have attempted to shed light on Kidney Disease prediction, the prophecy of Chronic Kidney Disease remains unsolved. For this reason, the main objective of this paper is to integrating some other machine learning algorithms like XGboost along with bayesian optimization for hyper parameter tuning of xgboost and improve CKD prediction along with a large feature set and strong non-linear dependencies within the data. The research derived a dataset from a sonograph showing a hospital from Karaikudi, Tamil Nadu from India. The dataset has 400 samples where 250 samples are positive for CKD and 150 samples are negative. This approach builds upon the previous work of Arumugham et al. (2023) who achieved a remarkable accuracy of 98.75% when using a deep neural network (DNN) model. The findings of this research offer insight into the use of advanced machine learning methods for the better prediction and management of chronic kidney disease.
Published
How to Cite
Issue
Section
FUDMA Journal of Sciences