Convolutional Neural Network (CNN) Quantum State Discrimination

Convolutional Neural Network-based quantum state discrimination as proposed by Ahmed et al. (2021) [1].

qsttoolkit.tomography.dlqst.CNN_classifier.architecture.build_classifier(data_input_shape: tuple) → Model[source]

Builds a CNN classifier network for QST data.

Parameters:

data_input_shape (tuple) – Shape of the input data.

Returns:

Classifier model.

Return type:

tf.keras.Model

class qsttoolkit.tomography.dlqst.CNN_classifier.model.CNNQuantumStateDiscrimination(X_train: ndarray, X_test: ndarray, y_train: ndarray, y_test: ndarray, label_encoder: LabelEncoder, early_stopping_patience: int = 5, lr_scheduler_factor: float = 0.5, lr_scheduler_patience: int = 3, dim=None)[source]

Bases: object

A class for training and evaluating a CNN classifier for quantum state discrimination.

X_train

Training data.

Type:

np.ndarray

X_test

Test data.

Type:

np.ndarray

y_train

Training labels.

Type:

np.ndarray

y_test

Test labels.

Type:

np.ndarray

label_encoder

Label encoder.

Type:

sklearn.preprocessing.LabelEncoder

early_stopping_patience

Number of epochs with no improvement after which training will be stopped. Defaults to 5.

Type:

int

lr_scheduler_factor

Factor by which the learning rate will be reduced. Defaults to 0.5.

Type:

float

lr_scheduler_patience

Number of epochs with no improvement after which learning rate will be reduced. Defaults to 3.

Type:

int

classify(X: ndarray) → ndarray[source]

Classifies a set of quantum states using the trained model.

Parameters:

X (np.ndarray) – Quantum states to classify.

Returns:

Predicted labels.

Return type:

np.ndarray

evaluate_classification(include_confusion_matrix: bool = True, include_classification_report: bool = True)[source]

Evaluates the model on the test data.

Parameters:

  • include_confusion_matrix (bool) – Whether to include the confusion matrix in the evaluation. Defaults to True.

  • include_classification_report (bool) – Whether to include the classification report in the evaluation. Defaults to True.

plot_training()[source]

Plots the training and validation accuracy and loss.

train(optimizer='adam', loss='sparse_categorical_crossentropy', metrics: list[str] = ['accuracy'], epochs: int = 20, batch_size: int = 32, validation_split: float = 0.2, verbose='auto')[source]

Compiles and trains the model.

Parameters:

  • optimizer – Optimizer to use in the training. Defaults to ‘adam’.

  • loss – Loss function to use in the training. Defaults to ‘sparse_categorical_crossentropy’.

  • metrics (list[str]) – Metrics to measure model performance during training. Defaults to [‘accuracy’].

  • epochs (int) – Number of epochs to train the model. Defaults to 20.

  • batch_size (int) – Training batch size. Defaults to 32.

  • validation_split (float) – Fraction of the training data to use as validation data. Defaults to 0.2.

  • verbose – Verbosity mode. Defaults to ‘auto’.