•  
  •  
 

Corresponding Author

Sourabh Shastri

Subject Area

Computer Science

Article Type

Original Study

Abstract

With the advancement of machine learning techniques, the introduction of the most accurate model has become a necessity. In real-world scenarios, every model has some constraints and assimilates errors, so their performance is not always highly efficient; this sparked the development of ensemble learning. The ensemble approach aims to consolidate the strengths of existing approaches and minimize their weaknesses or decision-making risks. The proposed diabetes prediction system encases a resampling filter, applied to balance the dataset and model builder method, i.e., without the SBV ensemble and with the SBV ensemble method. The model is initially built without using the SBV assembly method, and the evaluation of classifier output is done individually. Second, the proposed SBV ensemble framework is employed for building the model, i.e., stacking, bagging, and voting ensemble techniques are used with static base classifiers and dynamic meta classifiers having unique groups under each SBV category. The proposed methodology has been applied to two different diabetic datasets, including the Pima Indian Diabetes Dataset (Dataset 1) and the Early Stage Diabetes Risk Prediction Dataset (Dataset 2). The ensemble models have obtained higher acceptability in terms of accuracy and errors than single models. The best classifier from the SBV ensemble and the best SBV category group are compared to obtain the best model for diabetes detection. The findings of the present study reveal that the model's performance is sensitive to diabetes prediction and can also be used as a second opinion for a doctor in real-time.

Keywords

Ensemble, SBV framework, Base Classifiers, Meta Classifiers, Diabetes

Creative Commons License

Creative Commons Attribution 4.0 License
This work is licensed under a Creative Commons Attribution 4.0 License.

Download

Share

COinS