New publication by Dr. Denise Degen

 

Highlights

• Comparison of local and global sensitivity analysis demonstrating the need of global sensitivity analyses for robust model calibrations.
• Compensation for data sparsity through different weighting schemes.
• Physics-based machine learning approach to ensure the feasibility of the study.
• Robust sensitivity-driven model calibration to reliably identify model errors.
• Compensation of model errors through model calibration.