This Zenodo database supports the Imbalanced Regression Artificial Neural Network model for Substorms (IRANN-S), which applies weighting to different SuperMag AL (SML) index values. Contact Name: Xiangning Chu Email: chuxiangning@gmail.com Paper Title Imbalanced Regression Model of Auroral Electrojet Indices: Can We Predict Super Substorms? Authors & Affiliations Xiangning Chu – Laboratory for Atmospheric and Space Physics, University of Colorado Boulder, Boulder, CO, USA Lucas Jia – Department of Electrical & Computer Engineering, University of California Santa Barbara, Santa Barbara, CA, USA Robert L. McPherron – Department of Earth, Planetary, and Space Sciences, University of California, Los Angeles, CA, USA Xinlin Li – Department of Aerospace Engineering Sciences, University of Colorado Boulder, Boulder, CO, USA Jacob Bortnik – Department of Atmospheric and Oceanic Sciences, University of California, Los Angeles, CA, USA Rules of the Road The code and models are publicly distributed by the authors. To ensure continued research activity, users should contact the authors if these models and datasets are used in any publications, presentations, or any other formats. Authorship: If these products are used in a publication, please contact the authors and provide proper attribution. Usage: Data, plots, and derived products are provided under fair use limitations and can be redistributed. Repository Description This repository contains the Python code and model files used to develop the IRANN-S model. The models/ folder contains the trained model coefficients in H5 format, which can be loaded using TensorFlow. An example Python script (example.py) is provided to demonstrate model usage. Support For any questions, please feel free to contact the authors.