Comparative Analysis of Diabetes Diagnosis with Machine Learning Methods
Küçük Resim Yok
Tarih
2024
Yazarlar
Dergi Başlığı
Dergi ISSN
Cilt Başlığı
Yayıncı
Umut SARAY
Erişim Hakkı
info:eu-repo/semantics/openAccess
Özet
Diabetes is a disease that occurs when the body cannot regulate the level of sugar (glucose) in the blood. Early diagnosis of this disease is important in preventing more serious diseases that may arise later. Within the scope of this study, an attempt was made to optimize the diabetes data set for use by training it with different models. At the very beginning of the study, Logistic Regression, KNN, SVM (Support Vector Machine), CART (Classification and Regression Trees), RF (Random Forest), Adaboost, GBM (Gradient Boosting Machines), XGBoost (Extreme Gradient Boosting), LGBM (Light Gradient Boosting). Machine), CatBoost models were used. According to the results of the models, RF, LGBM, XGBoost accuracy, and f1 values were observed as the best models, respectively. As a result, in the Random Forest model, which produced the most successful results, Accuracy: 0.88, F1 Score: 0.84, and ROC AUC: 0.95 values were obtained, respectively.
Diabetes is a disease that occurs when the body cannot regulate the level of sugar (glucose) in the blood. Early diagnosis of this disease is important in preventing more serious diseases that may arise later. Within the scope of this study, an attempt was made to optimize the diabetes data set for use by training it with different models. At the very beginning of the study, Logistic Regression, KNN, SVM (Support Vector Machine), CART (Classification and Regression Trees), RF (Random Forest), Adaboost, GBM (Gradient Boosting Machines), XGBoost (Extreme Gradient Boosting), LGBM (Light Gradient Boosting). Machine), CatBoost models were used. According to the results of the models, RF, LGBM, XGBoost accuracy, and f1 values were observed as the best models, respectively. As a result, in the Random Forest model, which produced the most successful results, Accuracy: 0.88, F1 Score: 0.84, and ROC AUC: 0.95 values were obtained, respectively.
Diabetes is a disease that occurs when the body cannot regulate the level of sugar (glucose) in the blood. Early diagnosis of this disease is important in preventing more serious diseases that may arise later. Within the scope of this study, an attempt was made to optimize the diabetes data set for use by training it with different models. At the very beginning of the study, Logistic Regression, KNN, SVM (Support Vector Machine), CART (Classification and Regression Trees), RF (Random Forest), Adaboost, GBM (Gradient Boosting Machines), XGBoost (Extreme Gradient Boosting), LGBM (Light Gradient Boosting). Machine), CatBoost models were used. According to the results of the models, RF, LGBM, XGBoost accuracy, and f1 values were observed as the best models, respectively. As a result, in the Random Forest model, which produced the most successful results, Accuracy: 0.88, F1 Score: 0.84, and ROC AUC: 0.95 values were obtained, respectively.
Açıklama
Anahtar Kelimeler
Diabetes, Machine learning, Random forest
Kaynak
International Scientific and Vocational Studies Journal
WoS Q Değeri
Scopus Q Değeri
Cilt
8
Sayı
1
