Statistical Mechanics informed Neural Networks for Solar Wind modeling

This research project is the result of several approaches towards geomagnetic forecasting, all successful from their own way:

1. MHD informed Multi-Modal Neural Networks for Solar Wind modeling
2. Arasci: Attention mechanisms and Res-RNNs for Geomagnetic forecasting
   • https://github.com/Jorgedavyd/SMFGF-SpaceApps
   • https://github.com/Jorgedavyd/SMFGF_RNN
   • https://github.com/Jorgedavyd/Karai-AraSci
   • https://github.com/Jorgedavyd/Jakaira-AraSci
   • https://github.com/Jorgedavyd/Namandu-AraSci
   • https://github.com/Jorgedavyd/Colibri-AraSci
   • https://github.com/Jorgedavyd/Nande_Ru_Tenonde-AraSci
   • https://github.com/Jorgedavyd/Tupa-AraSci

Several tools were developed to ease the access to high quality scientific data and ML training:

1. Corkit: Democratizes and revamps calibration routines of corongraph data.
2. StarStream: Asynchronous data downlaoding for satellite data products.
3. LighTorch: A Pytorch and Lightning based framework for machine learning research.

Citation

@misc{starstream,
  author = {Jorge D. Enciso},
  title = {Statistical Mechanics informed Neural Networks for Solar Wind modeling},
  howpublished = {\url{https://github.com/Jorgedavyd/Vlasov-Poisson-PINNs-for-Solar-Wind-modeling}},
  year = {2024}
}