APPLICATION OF ARTIFICIAL NEURAL NETWORKS FOR ANALYSIS OF COSMIC RAY DATA

The analysis of particles that compose the galactic cosmic rays produce important data about the conditions of the interplanetary environment, improving the understanding of the Space Weather phenomena. The precision of this data, however, is directly correlated with our capacity to understand them. We used Multi-Layer Perceptrons model (MLP) and Nonlinear AutoRegressive with eXogenous inputs ( NARX) models of Artificial Neural Networks (ANN) to simulate the density of cosmic rays in the interplanetary medium, applying input vector data from the Advanced Composition Explorer (ACE) satellite located in the Lagrangian point (L1) about the Interplanetary Magnetic Field (IMF). The analysis of the model, topology, and characteristics of the most successful ANNs may not only indicated the best way to adjust this computational tool but also provide valuable clues about the relation between the galactic cosmic rays and the IMF data.