Classification using NFs

1. Install dependencies

pip install -r requirements.txt

2. Gaussianize the moon dataset

python train.py --config <type>

<type> can be selected from the config.yaml file (for e.g. master), which has all the hyperparameters.

Check the plots/sample directory for gaussinization and inverse function plots.

3. Sample output

• Gaussianized data

• Reconstructed data