AGDPM-Net: Hybrid Ensemble Framework for Genetic Disorder Prediction

Genetic disorders pose a significant challenge to human health, requiring early and precise diagnosis for effective treatment. This study introduces a hybrid machine learning framework that integrates multiple models to enhance prediction accuracy and reliability. A comprehensive dataset incorporating genetic, clinical, and demographic factors was utilized, with preprocessing techniques such as k-nearest neighbor (KNN) imputation for handling missing values, categorical encoding, and standardization. Three traditional machine learning models—random forest, gradient boosting, and support vector machine—were trained alongside deep learning models, including convolutional neural networks (CNN) and recurrent neural networks (RNN). A stacking ensemble approach was implemented, where predictions from these models served as input for a Random Forest meta-classifier, improving generalization and reducing biases. The ensemble model outperformed individual classifiers in accuracy, precision, recall, and F1-score, with CNN and RNN excelling in complex pattern recognition. This framework effectively processes diverse medical data types, combining structured tabular data analysis with deep learning’s ability to capture intricate relationships. Additionally, it addresses challenges such as data imbalance and missing values while reducing overfitting through ensemble learning. This study highlights the potential of AI driven methodologies in genetic disorder prediction, paving the way for scalable, automated, and more accurate diagnostic tools. Future research will focus on dataset expansion, refined feature selection, and advanced architectures like transformers and attention mechanisms to further enhance predictive performance and clinical applicability.