Hereditary-Based Disease Prediction Model using Machine Learning Techniques

Abstract

This research explored the use of machine learning for predicting hereditary diseases from genotype data. Both the Interpretable Genomic Neural Network (IGNN) and Random Forest models were trained and evaluated on genetic datasets. Results demonstrated that the Random Forest model achieved an overall accuracy of 86%, precision of 0.68, and recall of 0.52, while the IGNN model provided enhanced interpretability with comparable performance. Performance comparison showed that Random Forest outperformed baseline models such as Logistic Regression, highlighting its strength in predictive accuracy though still challenged by class imbalance. These achievements illustrate the potential of combining interpretable and ensemble learning approaches in early disease detection to support personalized medicine.