rachelgaoMIDS
commented
6 months ago - A visualization of the modeling pipeline (s) and subpipelines if necessary
- Families of input features and count per family
- Number of input features
- Hyperparameters and settings considered
- Loss function used (data loss and regularization parts) in latex
- Number of experiments conducted
- Experiment table with the following details per experiment:
- Baseline experiment
- Any additional experiments
- Final model tuned
- best results (1 to three) for all experiments you conducted with the following details
- Computational configuration used
- Wall time for each experiment