The datasets used in the demo are the following.
| Datasets | Ground Truth | Prediction |
|---|---|---|
| Training | sie-training.csv | pred |
| Validation | sie-validation.csv | pred |
| Testing | sie-testing.csv | pred |
The ground truth datasets were created from the following.
Dataset distribution: sie-dataset.toml
Problem configuration: sie-problem.toml
sie-dataset.csv which provides the lens parameters for simulation and calculation of roulette parameters.
sie-roulette.csv is the complete simulator output, including some columns ommitted in the data for machine learning.
training_log.csv gives the evolution log from training. This should be commented and analysed, but this has to wait for now.