Evolving Hybrid Cascade Neural Network Genetic Algorithm Space–Time Forecasting

Design: At the heart of time series forecasting, if nonlinear and nonstationary data are analyzed using traditional series, results will be biased. same time, just machine learning without any consideration given to input from not much information can obtained because model is a black box. Purpose: In order better study we extend combination propose hybrid cascade neural network considering metaheuristic optimization genetic algorithm in space–time forecasting. Finding: To further show utility algorithm, use various scenarios for training testing while also extending simulations by activation functions SoftMax, radbas, logsig, tribas on forecasting pollution data. During simulation, perform numerical metric evaluations root-mean-square error (RMSE), mean absolute (MAE), symmetric percentage (sMAPE) demonstrate that our models provide high accuracy speed up time-lapse computing.