LIME-HyGRU: An Explainable Hybrid Machine Learning and GRU Algorithm for Predicting Vector-Borne Diseases

Abstract

Vector-borne diseases remain a major global health concern, causing more than 700,000 deaths annually. These diseases are transmitted by vectors such as mosquitoes, ticks, flies, and fleas and are caused by a wide range of infectious agents, including bacteria, viruses, and parasites. Delayed diagnosis, particularly in regions with limited access to specialist healthcare results in increasing the complications and mortality. This leads the importance of developing a decision-support systems that can assist healthcare professionals in identifying these diseases at an early stage. This study proposes LIME-HyGRU, an explainable deep learning framework for the prediction of multiple vector-borne diseases by combining a Hybrid Gated Recurrent Unit (HyGRU) network with Local Interpretable Model-agnostic Explanations (LIME). This model enables both accurate disease prediction and transparent interpretation, allowing users to understand which clinical features contribute most to the final outcome. The combination of deep learning and explainable artificial intelligence (XAI) makes the model more suitable for clinical decision support. The framework is trained using labeled patient data and validated on previously unseen test samples to assess its predictive capability. Model performance is evaluated using widely accepted classification metrics, including accuracy, precision, recall, F1-score, confusion matrix, and receiver operating characteristic (ROC) analysis. The experimental findings indicate that LIME-HyGRU not only achieves reliable prediction performance but also offers meaningful explanations that improve the transparency and trustworthiness of the prediction process. These results demonstrate the potential of the proposed framework as an AI-assisted expert system for the early prediction and clinical management of vector-borne diseases.