LFGA–ADENN Hybrid Feature Extraction and Oversampling for Type 2 Diabetes Mellitus Detection in Mizoram
Science & Technology Journal · 27.05.2026 · 14 (Online First)
DOI: 10.22232/stj.2026.1 ↗Kurz gefasst
Fachartikel in der Zeitschrift Science & Technology Journal, online erschienen am 27.05.2026. Verfasst von Vanlalawmpuia Ralte und 6 weiteren Autor:innen. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0).
Themen: Diabetes mellitus
Abstract (Originalsprache)
Early detection of Type 2 Diabetes Mellitus using genomic characteristic is inspiring due to large dimensionality, huge missing values, and large number of class imbalance. Bioinformatics has made it easier to analyse complicated biological data, and the fact that an increasing number of this data is available has made machine learning techniques far more beneficial. The study proposes a hybrid system by integrating preprocessing, hybrid oversampling and feature extraction for the Mizoram population. Missing value filtering, K-Nearest Neighbours imputation, and categorical encoding are performed for curating data. The combination of the ADASYN and Edited Nearest Neighbors approach is used to address class imbalance. Feature extraction uses a genetic algorithm to choose useful representations after combining leaf indices from Decision Tree, Random Forest, and XGBoost models. Minority class detection is constantly improved by machine learning models and the hybrid LFGA–ADENN pipeline, reaching up to 0.921 F1-score and 0.920 MCC using XGBoost. LFGA–ADENN shows strong performance in imbalanced genomic T2DM prediction, improving recall while preserving precision when compared to no oversampling or single-method approaches.
Zitieren
APA
Ralte, V., Lalhmingliana, Kumar, N. S., Kumar, B. S., Lalrohlui, F., Zohmingthanga, J., & Vanlalhruaii (2026). LFGA–ADENN Hybrid Feature Extraction and Oversampling for Type 2 Diabetes Mellitus Detection in Mizoram. Science & Technology Journal, 14(Online First). https://doi.org/10.22232/stj.2026.1
Vancouver
Ralte V, Lalhmingliana, Kumar NS, Kumar BS, Lalrohlui F, Zohmingthanga J, et al. LFGA–ADENN Hybrid Feature Extraction and Oversampling for Type 2 Diabetes Mellitus Detection in Mizoram. Science & Technology Journal. 2026;14(Online First). doi:10.22232/stj.2026.1
BibTeX
@article{Ralte2026,
author = {Ralte, Vanlalawmpuia and Lalhmingliana and Kumar, Nachimuthu Senthil and Kumar, Brindha Senthil and Lalrohlui, Freda and Zohmingthanga, John and Vanlalhruaii},
title = {LFGA–ADENN Hybrid Feature Extraction and Oversampling for Type 2 Diabetes Mellitus Detection in Mizoram},
journal = {Science & Technology Journal},
year = {2026},
volume = {14},
number = {Online First},
doi = {10.22232/stj.2026.1}
}
EUVIMED