Abstract The effectiveness of machine learning models can be significantly averse to redundant and irrelevant features present in the large dataset which cause drastic performance degradation. This paper proposes IGRF-RFE: a hybrid feature selection method tasked for multi-class network anomalies using multilayer perceptron (MLP) network. IGRF-RFE exploits qualities both filter its speed wrappe...

1 Earth and Planetary Magnetism Group, Institut für Geophysik, Sonneggstrasse 5, ETH Zürich, CH-8092, Switzerland. 2NOAA/NGDC and CIRES, University of Colorado, U.S.A. 3British Geological Survey, Murchison House, West Mains Road, Edinburgh, EH9 3LA, U.K. 4Helmholtz Centre Potsdam, GFZ German Research centre for Geosciences, Telegrafenberg, 14473, Germany. 5School of Chemistry, University of New...

Relativistic particles entering the Earth’s magnetosphere, i.e. cosmic rays and solar energetic particles, are of prime space weather interest because they can affect satellite operations, communications, and the safety of astronauts and airline crews and passengers. In order to mitigate the hazards that originate from such particles one needs to predict the cutoff latitudes of such particles a...

Abstract The International Geomagnetic Reference Field (IGRF) is a multi-institute model of the Earth’s magnetic field, compactly described by sets up to 195 spherical harmonic (Gauss) coefficients degree and order 13, which allows continuous evaluation field at any location time on or above surface. It developed from satellite ground-based magnetometer data describes large-scale variation in s...

We present a statistical study of magnetic field vector residual between Swarm observations and two Earth models: the 13th generation International Geomagnetic Reference Field (IGRF) model CHAOS-7 model. Statistics these residuals are important for estimating potential errors satellite operations when using models as reference, well magnetosphere-ionosphere-thermosphere studies examining energy...