ARCHIVES
VOL. 10, ISSUE 1 (2026)
A novel deep ensemble learning model for medical disease diagnosis: design, evaluation, and performance analysis
Authors
Kulkarni Usha Bhimrao, Dr. Sanjay Kumar
Abstract
The early and correct diagnosis of chronic diseases
is one of the most important problems of contemporary healthcare because of the
complexity and heterogeneity of medical data. This paper presents a new deep ensemble
learning architecture in medical disease diagnosis that combines both advanced preprocessing,
feature selection and classification features to enhance predictive performance.
The model uses Multivariate Imputation by Chained Equations (MICE) to deal with
missing values and a Synergetic Outlier Factor (SOF)-based approach to robust outlier
detection. Gini Importance and Permutation Importance approaches are used to find
the most informative attributes. The refined data is then utilized in training an
ensemble of machine learning models, including XGBoost, Bagging, and Multi-Layer
Perceptron, combined by a hard voting strategy. Moreover, calibration-boosted deep
ensemble model (cbForest) is included to improve the classification accuracy and
generalization. The proposed framework will be tested on several medical datasets,
including ILPD, PIDD and Diabetes datasets. The experimental evidence shows that
the suggested model is better in comparison with the traditional machine learning
models, as it is more accurate and robust. These results emphasize the usefulness
of the suggested method as a well-grounded tool of intelligent medical diagnosis.
Download
Pages:68-75
How to cite this article:
Kulkarni Usha Bhimrao, Dr. Sanjay Kumar "A novel deep ensemble learning model for medical disease diagnosis: design, evaluation, and performance analysis". International Journal of Advanced Engineering and Technology, Vol 10, Issue 1, 2026, Pages 68-75
Download Author Certificate
Please enter the email address corresponding to this article submission to download your certificate.