Customized and Automated Machine Learning-Based Models for Diabetes Type 2 Classification.

Area of Science:

• Medical Informatics
• Computational Biology
• Machine Learning in Healthcare

Background:

• Accurate classification of type 2 diabetes is crucial for effective patient management.
• Existing classification methods may benefit from advanced machine learning approaches.
• The Practice Fusion dataset offers a valuable resource for clinical data analysis.

Purpose of the Study:

• To develop and compare machine learning models for classifying type 2 diabetes patients.
• To evaluate the performance of Random Forest, Support Vector Classifier, AdaBoost, ensemble, and automated machine learning models.
• To identify the most effective model for type 2 diabetes patient classification.

Main Methods:

• Utilized the Practice Fusion dataset for model development.
• Implemented Random Forest (RF), Support Vector Classifier (SVC), AdaBoost, ensemble, and automated machine learning (AutoML) models.
• Employed a five-fold cross-validation scheme for robust performance evaluation.

Main Results:

• The automated machine learning (AutoML) model demonstrated superior performance compared to individual and ensemble models.
• AutoML achieved the highest scores across all four evaluation measures used in the study.
• All implemented models showed varying degrees of classification accuracy.

Conclusions:

• Automated machine learning (AutoML) presents a highly effective approach for type 2 diabetes patient classification.
• The findings suggest that AutoML can enhance diagnostic accuracy in clinical settings.
• Further research can explore AutoML's potential in predicting diabetes progression and complications.