ECCE-2025-data

Data Related to ECCE 2025 paper on coreloss prediction

Structure

• Resampled
  • Sample data with 24 points. (Ready to use as examples or training)
• magnet-challenge-version
  • copy ASU model
• src
  • inference
    • excel
    • matlab
    • pytorch
    • vanilla web
    • magnet-princeton (github linked)
    • magnethub (github linked)
    • magnet-engine (github linked)
    • google sheet -- url
  • training
    • ecce train :(deployed at) https://ecce-2025-data.streamlit.app/
    • magnet challenge train
  • main.py

Related links

• ECCE-2025-data

• magnet fork

• magnet-engine fork

• ECCE-2025-code

• MagNetChallenge2023 fork

Citation

Cite the work as:

E. Havugimana and M. K. Ranjram, “A Simple Data-Driven Machine Learning-based Software Package for Accurate Magnetic Core Loss Computation,” in 2025 IEEE Energy Conversion Congress and Exposition (ECCE), 2025.