2024 · 14 citations · 8 references
ComputingBiomedical InformaticsHealth InformaticsClinical Decision SupportData MiningMachine LearningClassificationComputer-aided DiagnosisDecision Support In HealthcareDecision TreesDiagnostic Decision SupportRenal DiseaseClassification AccuracyDiagnostic AlgorithmsDiagnostic Prediction ModelsExtreme Gradient BoostingNaive BayesNaive Bayes ClassificationRandom Forest ClassificationXgboost
This research presents a novel diagnostic method for determining the severity of kidney disease by combining fuzzy logic with the Extreme Gradient Boosting (XGBoost) algorithm. The model makes use of patient data from the UCI ML Repository, which includes clinical measures and demographic data. Before the XGBoost algorithm receives the data, fuzzy logic preprocesses it, turning unclear inputs into organized information. With an accuracy of 93.65%, the Fuzzy-Enhanced XGBoost model outperforms the SVM, Random Forest, Decision Tree, and Naive Bayes algorithms as well as standalone XGBoost. The incorporation of fuzzy logic enhances the model’s capacity to manage imprecise and noisy data, resulting in enhanced robustness and generalization. This study improves interpretability and reliability by addressing the shortcomings of conventional methods in managing ambiguous medical data. A comparative examination demonstrates that the Fuzzy-Enhanced XGBoost model provides superior management of uncertainty in addition to increasing classification accuracy. To sum up, this model offers a powerful and understandable diagnostic tool that represents considerable progress in the classification of renal disease severity.
8
Suyuan Peng, Feichen Shen, Andrew Wen et al. · Journal of Medical Internet Research · 2019 · 28 citations · Full text