Sunspot prediction problem based on PSO-RBF

Sunspots, as an important indicator of solar activity, reflect changes in solar activity. In recent years, many scientists have made use of advanced observation techniques and data models to predict the number of sunspots. Starting from the vision of data analysis, based on the number and area data of sunspots, we build an RBF neural network model that uses radial basis functions to map the input space nonlinearly in order to capture complex relationships in the data. In order to improve the prediction accuracy of the model, PSO algorithm was introduced to optimize the model parameters. Finally, we predict the number and area of sunspots in the current and next solar cycles, the specific results are shown in the text, and the reliability of the model is reasonably explained by the RMSE index.