Predictive Modeling of Global Vector-Borne Diseases: Leveraging Machine Learning for Intervention Strategies

This paper addresses the evolving landscape of global infectious diseases, emphasizing the pivotal role of vector-borne diseases in public health. Focusing on eleven formidable diseases, including Chikungunya, Dengue, Zika, Yellow Fever, West Nile Fever, Malaria, Rift Valley Fever, Tungiasis, Japanese Encephalitis, Plague, and Lyme Disease, the study utilizes machine learning algorithms for prediction, risk factor identification, and correlation analysis. Data, collected from diverse sources, undergoes thorough preprocessing, ensuring quality and consistency. Four machine learning models are employed-Multilayer Perceptron, Sequential Minimal Optimization, J48 Decision Tree, and Naive Bayes-revealing the Naive Bayes model as particularly promising with a 94% correlation coefficient and high precision. Association rules shed light on complex attribute relationships, guiding future research, and interventions. The ranked attributes, led by itchiness, provide valuable insights for understanding and mitigating vector-borne diseases. This research contributes to a deeper comprehension of these diseases, facilitating informed decision-making and effective public health strategies amid evolving global health challenges.