Forecasting and comparing between five moderate geomagnetic storms in 2022 using artificial neural networks

The current work investigates the precursor that follows coronal mass ejection (CME), through studying five moderate geomagnetic storms in 2022. We have employed the Artificial Neural Network (ANN) and supervised machine learning models to predict the SYM-H for geomagnetic storms during the solar cycle. The estimation yielded satisfactory accuracy including mean absolute error (MAE), mean square error (MSE), root mean square error (RMSE), and correlation coefficient (R ² ). To demonstrate the method's robustness, we have compared the predicted data set with real-world data and evaluated its performance against other supervised machine learning algorithms for regression problems, namely Decision Tree Regressor, Gradient Boosting Regressor, AdaBoost Regressor, and Linear Regression. Results revealed the proficiency of the ANN as an effective predicting tool over the Supervised machine learning for the SYM-H index.