Eine Plattform für die Wissenschaft: Bauingenieurwesen, Architektur und Urbanistik
Eine Plattform für die Wissenschaft: Bauingenieurwesen, Architektur und Urbanistik
Predicting Pre-Diabetic and Diabetes in Adults and Elderlies Using Machine Learning
Diabetes, a chronic metabolic disorder charac-terized by elevated blood glucose levels, poses a significant global health challenge. This paper aims to utilize machine learning algorithms for early prediction and diagnosis of diabetes. Mendeley Diabetes dataset, consisting of 1000 instances, was used in this study. The dataset has 11 input attributes and a dependent variable that classifies if a person has diabetes, is pre-diabetic, or healthy. The dataset is imbalanced. Five machine learning algorithms have been used in this study. The classification algorithms are Random Forest (RF), Decision Tree (DT), K-Nearest Neighbors (KNN), Support Vector Machine (SVM), and Naïve Bayes (NB). Several experiments have been done to enhance and get the best accuracy. After applying feature selection, we found that six features out of eleven are enough to predict pre-diabetes and diabetes. Moreover, to ensure that the high accuracy is not biased toward the large class, SMOTE was applied to the training set to balance the dataset. Among all the experiments, the best model generated was using RF classifier with 99.7% accuracy, after feature selection and balancing the training dataset.
Predicting Pre-Diabetic and Diabetes in Adults and Elderlies Using Machine Learning
Diabetes, a chronic metabolic disorder charac-terized by elevated blood glucose levels, poses a significant global health challenge. This paper aims to utilize machine learning algorithms for early prediction and diagnosis of diabetes. Mendeley Diabetes dataset, consisting of 1000 instances, was used in this study. The dataset has 11 input attributes and a dependent variable that classifies if a person has diabetes, is pre-diabetic, or healthy. The dataset is imbalanced. Five machine learning algorithms have been used in this study. The classification algorithms are Random Forest (RF), Decision Tree (DT), K-Nearest Neighbors (KNN), Support Vector Machine (SVM), and Naïve Bayes (NB). Several experiments have been done to enhance and get the best accuracy. After applying feature selection, we found that six features out of eleven are enough to predict pre-diabetes and diabetes. Moreover, to ensure that the high accuracy is not biased toward the large class, SMOTE was applied to the training set to balance the dataset. Among all the experiments, the best model generated was using RF classifier with 99.7% accuracy, after feature selection and balancing the training dataset.
Predicting Pre-Diabetic and Diabetes in Adults and Elderlies Using Machine Learning
Rayes, Lubana Al (Autor:in) / Haggag, Menatalla (Autor:in) / Afyouni, Imad (Autor:in)
03.06.2024
361704 byte
Aufsatz (Konferenz)
Elektronische Ressource
Englisch
Predicting diabetic retinopathy based on routine laboratory tests by machine learning algorithms
| Springer Verlag | 2025
| DOAJ | 2024
Predicting IRI Using Machine Learning Techniques
| Springer Verlag | 2023
| DOAJ | 2019