Data preparation

Functions that prepare data for the training process.

dataset_loader()

Creates datasets and dataloaders

prepare_to_adv()

Prepares data for adversarial model

preprocess()

Preprocesses data for training

Training workflow

Functions that are directly involved in training process.

fair_train()

Trains fairly the PAN model

pretrain()

Pretrains both classifier and adversarial

create_model()

Creates a simple neural network model

pretrain_net()

Pretrains the neural network

Outcome preparation

Functions that prepare the outcomes, measures performance and plot it.

make_preds()

Makes binary predictions

make_preds_prob()

Makes probabilistic predictions

calc_STP()

Calculates STP ratio

eval_accuracy()

Calculates accuracy

explain_pan()

Creates explainer of PAN model

plot_monitor()

Plots monitored data