Fig. 5 | Scientific Reports

Fig. 5

From: Forecasting solar energetic particles using multi-source data from solar flares, CMEs, and radio bursts with machine learning approaches

Fig. 5

Feature importance scores for the RF model applied to the fixed frequency dataset. The x-axis lists the features used in the model, including solar flare characteristics, CME characteristics, and fixed frequency radio burst characteristics. The y-axis displays the importance score of each feature, reflecting its relative impact on the model’s predictions. Notably, linear speed, angular width, and second order speed of CME emerge as the most significant features, contributing the most to the model’s predictive power.

Back to article page