From Monte Carlo to Sensitivity Analysis and Metamodels

Upload input-output datasets from Monte Carlo simulations and perform sensitivity analysis and construct metamodels

Sensitivity analysis with no sampling requirements

The applied sensitivity analysis, TOM, does not require a specific sampling strategy.
Get visual overview on how much the model inputs affect each output parameter.

Use metamodels to estimate input space more thoroughly

Train neural networks for individual, or all, outputs. These fast metamodels can be make additional predictions within the multi-dimensional input space. Or used to run optimization.

Work online, use Google drive, or run locally

Depending on your Python experience and privacy concerns, you have different options to use this notebook.

1. Run the Github Notebook online using Colab
2. Run using Colab with access to your local Google Drive account.
3. Run locally using your own Python interpreter and Jupyter Lab