Research Article
Predictive Modeling for Women Anemia Diagnosis Using Hybrid Machine Learning Approaches
@INPROCEEDINGS{10.4108/eai.19-12-2025.2362551, author={Gagandeep Kaur and Ashima Kukkar and Bhumika Chopra and Armanpreet Kaur}, title={Predictive Modeling for Women Anemia Diagnosis Using Hybrid Machine Learning Approaches}, proceedings={Proceedings of the 1st International Conference on Computational Techniques in Data Science, IICTDS 2025, 19-20 December 2025, Chandigarh, India}, publisher={EAI}, proceedings_a={IICTDS}, year={2026}, month={11}, keywords={Anemia Prediction Machine Learning Ensemble Model XGBoost Random Forest Women’s Health Classification}, doi={10.4108/eai.19-12-2025.2362551} }- Gagandeep Kaur
Ashima Kukkar
Bhumika Chopra
Armanpreet Kaur
Year: 2026
Predictive Modeling for Women Anemia Diagnosis Using Hybrid Machine Learning Approaches
IICTDS
EAI
DOI: 10.4108/eai.19-12-2025.2362551
Abstract
Anemia remains one of the most prevalent health concerns among women, significantly impacting maternal and overall well-being, particularly in developing regions. Early and accurate diagnosis is essential for timely medical intervention. This study proposes a robust machine learning-based framework to predict anemia in women using clinical and demographic features. The dataset, sourced from Kaggle, underwent comprehensive preprocessing, including feature encoding, normalization, and outlier handling. Multiple classification algorithms—Logistic Regression, Decision Tree, Random Forest, Support Vector Machine, XGBoost, LightGBM, Gradient Boosting, AdaBoost, K-Nearest Neighbors, and Naïve Bayes—were implemented and evaluated based on accuracy, precision, recall, F1-score, and AUC-ROC.Experimental findings reveal that ensemble-based algorithms outperform traditional classifiers, with XGBoost and Random Forest achieving accuracies of 93.61% and 92.08%, respectively. To enhance predictive performance, a hybrid ensemble model (XGB–RF) integrating the strengths of both models was developed using a weighted voting approach. The proposed hybrid achieved a peak accuracy of 95.42%, F1-score of 0.94, and AUC-ROC of 0.96, outperforming all individual models. The results substantiate the potential of hybrid ensemble learning in medical diagnosis systems, providing a reliable and scalable decision-support mechanism for anemia screening among women.