An Innovative Model of Spare Attention-Based Elman RNN and Optimized Micro-Level Selection with Parameter Analysis for Dengue Outbreak Prediction

For the past few years, the Dengue Fever (DF) becomes an inevitable disease caused by mosquito spread over the region and it affects around 4 billion individuals across globally. Due to its rapid spreading nature, the more people gets affected, mainly the blood cells, which could early lead to death as well. Thus, it increases the mortality and morbidity rate as it outbreaks with huge damage. In order to account of that, the forecasting process of dengue outbreak is necessary to reduce the mosquito spread by the computation of outbreak level. Therefore, an intelligent approach is proposed for controlling and predicting the dengue outbreak. Firstly, the required input sources are fetched from the benchmark datasets. Subsequently, the optimal set of features are selected or chosen using Statistical Adaptive Hippopotamus Optimization (St-AHO). Finally, the resultant optimal features are fed into the Sparse Attention with Elman Recurrent Neural Network (SA-ERNN). In order to do the better prediction, the parameter analysis is done by using St-AHO, where the micro-level values for features are adjusted to predict the dengue outbreak. Finally, the experimental analysis is done using distinct metrics and compared with other existing approaches. Hence, the better results achieve to demonstrate the system significance while predict the dengue outbreak.