Neural network classification of eigenmodes in the magnetohydrodynamic spectroscopy code Legolas

To predict the immediate evolution of a plasma system, one needs to identify the nature of the dominant instability. In this work, a neural network is employed to address a non-binary classification problem of instabilities in astrophysical jets, whose natural oscillations and instabilities are quantified with the magnetohydrodynamic spectroscopy code Legolas. The trained models exhibit reliable performance in the identification of the two instability types supported by these jets. To improve the neural network aided classification process, techniques for training data augmentation and refinement of predictions for general eigenproblems are discussed.