microquantum.qml.classifier

Variational quantum classifier for quantum machine learning.

Implements a variational quantum classifier that combines a data encoding feature map with a parameterized variational circuit for binary and multi-class classification.

Module Contents

class microquantum.qml.classifier.ClassifierResult

Bases: microquantum._json.JSONSerializable

Result from a variational classifier.

Variables:

• predictions – Predicted class labels for each input.

• probabilities – Class probabilities for each input.

• accuracy – Classification accuracy (if true labels provided).

• optimizer_result – Underlying VQE optimization result.

• num_classes – Number of output classes.

predictions: list[int] = []

probabilities: list[list[float]] = []

accuracy: float = 0.0

optimizer_result: microquantum.optimizers.base.OptimizerResult | None = None

num_classes: int = 2

class microquantum.qml.classifier.VariationalClassifier(num_features, num_classes=2, encoder=None, ansatz_depth=2, optimizer=None)

Variational quantum classifier.

Combines a data encoding feature map with a parameterized ansatz for classification. The classifier is trained by minimizing a cross-entropy loss on labeled data.

For binary classification, the output is a single qubit measured in the Z-basis. For multi-class, multiple output qubits are used.

Parameters:

• num_features (int) – Number of input features.

• num_classes (int) – Number of output classes (default 2).

• encoder (Optional[microquantum.qml.encoding.BaseEncoder]) – Feature map for data encoding. If None, uses AngleEncoding.

• ansatz_depth (int) – Depth of the parameterized ansatz circuit.

• optimizer (Optional[microquantum.optimizers.base.Optimizer]) – Classical optimizer for training.

property num_features: int

Return type:

int

property num_classes: int

Return type:

int

property num_qubits: int

Return type:

int

property parameters: dict[microquantum.core.parameter.Parameter, float]

Return type:

dict[microquantum.core.parameter.Parameter, float]

predict(X, params=None)

Predict class labels for a set of inputs.

Parameters:

• X (list[list[float]]) – Feature vectors to classify.

• params (Optional[dict[microquantum.core.parameter.Parameter, float]]) – Optional parameter override. Uses self.params if None.

Returns:

ClassifierResult with predictions and probabilities.

Return type:

ClassifierResult

score(X, y, params=None)

Compute classification accuracy.

Parameters:

• X (list[list[float]]) – Feature vectors.

• y (list[int]) – True class labels.

• params (Optional[dict[microquantum.core.parameter.Parameter, float]]) – Optional parameter override.

Returns:

Accuracy between 0 and 1.

Return type:

float