Dengue Fever Prediction Empowered by Radial Basis Function Networks, Dynamic Mode Decomposition, and Learning-Based Foraging Algorithm

Abstract

Dengue fever is presently considered a major health threat that must be addressed. External and internal factors that induce nonlinear oscillations in the occurrence of dengue disease have made optimal resource allocation stimulating. Public health initiatives aimed at eradicating this vector use accurate and timely data through field processes. In dengue-endemic nations, early dengue prediction remains a main concern for public health. Developing a robust forecast model for accurate dengue prediction is a challenging task that can be accomplished through the application of a diversity of data modeling methods. Dengue fever remains a significant global public health concern, and the effort and dynamism make traditional methods challenging to predict. Combining Radial Basis Function Networks (RBFNs) with Dynamic Mode Decomposition (DMD) and Learning-Based Foraging Algorithm (LBFA), a novel method is applied for predicting dengue disease. By including RBFNs, a robust machine learning tool, the non-linear associations and patterns are extracted from dengue fever data. DMD, a data-driven decay method, enables the elucidation of the important modes and dynamics in time series data, thereby providing valuable insights into disease communication patterns. Also, LBFA, inspired by the foraging behavior of fruit flies, develops the parameters of the RBFN model, thereby improving its precision and flexibility. Using historical dengue fever data from multiple regions, inclusive tests were conducted to evaluate our proposed RBFN-DMD method. In terms of accuracy and reliability, the RBFN-DMD with LBFA outperforms predictive forecasting methods in predicting dengue fever. Additionally, our method provides interpretable insights into the driving forces and dynamics of dengue fever transmission, enabling public health establishments to make more cognizant decisions about disease prevention and control.