Efficient time-series forecasting of nuclear reactions using swarm intelligence algorithms

In this research paper, we focused on the developing a secure and efficient time-series forecasting of nuclear reactions using swarm intelligence (SI) algorithm. Nuclear radioactive management time series for casting is problem to be addressed if power deliver major part our energy consumption. This explains how SI processing techniques can used automate accurate reaction forecasting. The goal study was use analysis understand patterns in dataset while intelligence. results obtained by training algorithm longer periods predicting events with 94.58 percent accuracy, which higher than deep convolution neural networks (DCNNs) 93% accuracy all predictions, such as number active reactions, see improve. Our earliest determining best settings preprocessing working certain reaction, fusion task: took 0-500 ticks being trained 300 epochs