Train

The mylib.train contains classes:

• mylib.train.Trainer

The mylib.train contains functions:

• mylib.train.cv_parameters()

class mylib.train.SyntheticBernuliDataset(n=10, m=100, seed=42)

Base class for synthetic dataset.

class mylib.train.Trainer(model, X, Y, seed=42)

Base class for all trainer.

eval(output_dict=False)

Evaluate model for initial validadtion dataset.

test(X, Y, output_dict=False)

Evaluate model for given dataset.

Parameters:

• X (numpy.array) – The array of shape num_elements \(\times\) num_feature.

• Y (numpy.array) – The array of shape num_elements \(\times\) num_answers.

train()

Train model

mylib.train.cv_parameters(X, Y, seed=42, minimal=0.1, maximum=25, count=100)

Function for the experiment with different regularisation parameters

and return accuracy and weidth for LogisticRegression for each parameter.

Parameters:

• X (numpy.array) – The array of shape num_elements \(\times\) num_feature.

• Y (numpy.array) – The array of shape num_elements \(\times\) num_answers.

• seed (int) – Seed for random state.

• minimal (int) – Minimum value for the Cs linspace.

• maximum (int) – Maximum value for the Cs linspace.

• count (int) – Number of the Cs points.