Jul 2026· International Journal of Science and Environment (IJSE)· 0 citations· 25 references
TL;DR
The experimental results show that for the GNB model, the best performance was achieved using the combination of StandardScaler, SMOTE, and SelectKBest (k=5), reaching an accuracy of 94.53%, precision 98.36%, recall 90.91%, and f1-score 94.49%.
Abstract
The significant increase in diabetes mellitus cases within the community demands a technology-based solution that can provide accurate, efficient, and reliable predictions. This study aims to evaluate the impact of various data preprocessing schemes on the performance of the Gaussian Naive Bayes (GNB) and Support Vector Machine (SVM) algorithms in classifying diabetes risk. The dataset used in this research was sourced from the UCI Machine Learning Repository and consists of 520 records with 16 symptom features and 1 target label. The preprocessing stages include handling missing values, encoding categorical features, normalizing numerical data using StandardScaler, balancing the dataset with the Synthetic Minority Over-sampling Technique (SMOTE), and feature selection using the SelectKBest method. A total of nine preprocessing scheme combinations were tested for each algorithm. The experimental results show that for the GNB model, the best performance was achieved using the combination of StandardScaler, SMOTE, and SelectKBest (k=5), reaching an accuracy of 94.53%, precision 98.36%, recall 90.91%, and f1-score 94.49%. Meanwhile, for the SVM model, the highest performance was obtained through the combination of StandardScaler and RBF kernel hyperparameter tuning, achieving an accuracy of 99.04%, precision 99.05%, recall 99.04%, and f1-score 99.03%. The evaluation was conducted using metrics such as accuracy, precision, recall, F1-score, confusion matrix, and learning curve visualization. These findings highlight the critical role of proper preprocessing in enhancing predictive model performance. This study is expected to serve as a reference for developing early detection systems for diabetes based on machine learning.
Diabetes is a chronic metabolic disease characterized by high blood glucose levels that can lead to severe complications affecting various organs. The increasing prevalence of diabetes requires an accurate and efficient diagnostic approach to support early detection and medical decision-making. The development of machine learning technology provides opportunities to assist disease diagnosis by analyzing patient health data and identifying patterns associated with diabetes conditions. This study aims to compare the accuracy performance of the Naive Bayes and CatBoost algorithms in diagnosing diabetes using the confusion matrix evaluation method. The research method consists of several stages, including data collection, data preprocessing, model selection, model training, and model evaluation. The dataset used was obtained from Kaggle, consisting of 768 records with nine attributes, including pregnancies, glucose level, blood pressure, skin thickness, insulin, body mass index, diabetes pedigree function, age, and outcome classification. The models were trained using 80% training data and 20% testing data. The evaluation results showed that the Naive Bayes algorithm achieved an accuracy of 79%, while the CatBoost algorithm obtained a higher accuracy of 81%. These findings indicate that CatBoost provides better classification performance compared to Naive Bayes for diabetes diagnosis in this dataset. Therefore, CatBoost can be considered a more effective machine learning approach to support diabetes prediction. Future research is recommended to utilize larger datasets, optimize model parameters, and explore other advanced algorithms to improve diagnostic accuracy.
Basyach Aryodito Tranggono Heryanto, Onny Marleen· Journal of social research· 0 citations
Abstract. Early prediction of type 2 diabetes mellitus is important to facilitate faster and more accurate diagnosis and clinical decision-making. This study aims to compare the performance of the Random Forest and Support Vector Machine (SVM) algorithms in predicting diabetes and to analyze the effect of applying the Synthetic Minority Over-sampling Technique (SMOTE) to imbalanced data. The study used the Bangladesh Diabetes 2025 dataset, following the stages of data selection, preprocessing, data transformation, modeling, and evaluation. The preprocessing stage included median imputation and the removal of duplicate data, while feature standardization was performed after data splitting to prevent data leakage. The dataset was split using an 80:20 ratio with a stratification technique. The study applied two experimental scenarios: one without SMOTE and one with SMOTE applied only to the training data. Model evaluation was conducted using the metrics accuracy, precision, recall, and F1-score based on the test data. The results show that Random Forest outperforms SVM. In the scenario without SMOTE, Random Forest achieved an accuracy of 94.8%, precision of 96.4%, recall of 97.0%, and an F1-score of 96.7%, while SVM achieved an accuracy of 91.5% and an F1-score of 94.6%. After applying SMOTE, Random Forest’s performance improved slightly to an accuracy of 95.3% and an F1-score of 97.0%, while SVM’s performance declined to an accuracy of 90.6% and an F1-score of 93.9%. The results of the study show that Random Forest is the best model for predicting type 2 diabetes mellitus in the dataset used.
Baharudin Yusuf· Jurnal Informatika dan Tekni...· 0 citations
Diabetes mellitus is a chronic metabolic disorder whose global prevalence continues to rise, creating an urgent need for scalable, low-cost tools for early risk identification. This study evaluates the effectiveness of five machine learning algorithms (Random Forest, XGBoost, Support Vector Machine [SVM], CatBoost, and TabNet) for predicting diabetes risk from routinely available clinical and lifestyle variables. Using the Pima Indians Diabetes dataset, a preprocessing pipeline was applied that included median imputation of physiologically implausible zero values, standardized (Z-score) feature scaling, Boruta-based feature selection, and class-imbalance handling through class-weight adjustment and the Synthetic Minority Oversampling Technique (SMOTE). Models were trained on an 80/20 train-test split and assessed using accuracy, precision, recall, F1-score, and the area under the receiver operating characteristic curve (ROC-AUC). Random Forest achieved the strongest overall performance (accuracy 0.753; F1-score 0.689; ROC-AUC 0.810), followed by CatBoost, SVM, and XGBoost, whereas TabNet performed worst with very low recall for the diabetic class. The best-performing model (Random Forest) was deployed in a lightweight Flask web application that returns a probability-based diabetes risk assessment, categorising each prediction as low, moderate, or high risk together with a tailored recommendation. The findings confirm that ensemble tree-based methods, particularly Random Forest, provide a reliable, interpretable, and deployable basis for diabetes risk screening, especially in resource-constrained settings. Key limitations include dataset homogeneity, residual class imbalance, and limited feature coverage.
T. Olayinka· FUDMA Journal of Sciences· 0 citations
The study comes to the conclusion that headline accuracy is an unreliable guide in imbalanced medical prediction, that imbalance handling can change a model's practical usefulness, and that this benefit is strongly algorithm-dependent, meaning that the decision to resample should be based on the algorithm and the screening priorities rather than being applied consistently.
A. Oduroye, Temilade Opanuga, Esther Tosin Akanbi et al.· International journal of re...· 0 citations
In this study, we aimed to create a system that uses machine learning to detect and classify diabetes in an e-healthcare setting. We used Ensemble Decision Tree algorithms for selecting important features from a large set of data. Detecting diabetes accurately is a big challenge for researchers, especially in e-healthcare environments. Many existing systems have problems like slow processing and low accuracy. To fix these issues, we built a diabetes diagnosis system that includes data cleaning, feature selection, and classification. We tested the system using methods to check its effectiveness. We used a filter method based on the Decision Tree algorithm to choose the most important features. We also used two types of ensemble learning Decision Tree algorithms, Ada Boost and Random Forest, for feature selection and compared their performance with wrapper-based methods. The Decision Tree classifier was used to separate healthy individuals from those with diabetes. The results showed that using the selected features improved the model's classification performance and achieved the best accuracy. The system also performed better than previous methods due to the combination of different feature sets.
Twarita Taru, Madanika G· International Research Journ...· 0 citations
The results demonstrate that ensemble learning structures, which combine the strengths of multiple models, can provide more reliable decision support in critical areas such as healthcare and emphasise that ensemble learning structures, which combine the strengths of multiple models, can provide more reliable decision support in critical areas such as healthcare.
Yasin Karakuş, Pınar Özen· Journal of Innovative Engine...· 0 citations
