uniqc.algorithms.core.training.qcnn module

Quantum Convolutional Neural Network (QCNN) with TorchQuantum backend.

Implements convolutional and pooling layers on qubits for quantum state classification, using TorchQuantum’s native PyTorch autograd.

class uniqc.algorithms.core.training.qcnn.QCNNClassifier(n_qubits=8, n_classes=2)[source]

Bases: Module

Quantum Convolutional Neural Network for classification.

Applies convolutional and pooling layers that progressively reduce the number of active qubits, then measures the final qubit.

Parameters:

  • n_qubits – Number of qubits (preferably power of 2).

  • n_classes – Number of output classes.

forward(x=None)[source]

Classify quantum state.

Returns:

Probability-like output in [0, 1].