Add TrainableFidelityStatevectorKernel

[declanmillar]

Summary

Adds a TrainableFidelityStatevectorKernel class. Closes #638.

Details and comments

For consistency with TrainableFidelityQuantumKernel, an equivalent class that inherits from FidelityStatevectorKernel and TrainableKernel has been added.

To avoid some code repetition, _parameter_array has been moved to TrainableKernel and this now raises the error that was previously in evaluate of TrainableFidelityQuantumKernel.

TrainableFidelityStatevectorKernel is now unit tested alongside TrainableFidelityQuantumKernel and also with QuantumKernelTrainer.

Performance

# External imports
from sklearn import metrics
import numpy as np
from timeit import default_timer

# Qiskit imports
from qiskit import QuantumCircuit
from qiskit.circuit import ParameterVector
from qiskit.algorithms.optimizers import SPSA
from qiskit.circuit.library import ZZFeatureMap
from qiskit.utils import algorithm_globals
from qiskit_machine_learning.kernels import TrainableFidelityQuantumKernel, TrainableFidelityStatevectorKernel
from qiskit_machine_learning.kernels.algorithms import QuantumKernelTrainer
from qiskit_machine_learning.algorithms import QSVC
from qiskit_machine_learning.datasets import ad_hoc_data

algorithm_globals.random_seed = 12345

adhoc_dimension = 2
X_train, y_train, X_test, y_test, adhoc_total = ad_hoc_data(
    training_size=20,
    test_size=5,
    n=adhoc_dimension,
    gap=0.3,
    plot_data=False,
    one_hot=False,
    include_sample_total=True,
)

# Create a rotational layer to train. We will rotate each qubit the same amount.
training_params = ParameterVector("θ", 1)
fm0 = QuantumCircuit(2)
fm0.ry(training_params[0], 0)
fm0.ry(training_params[0], 1)

# Use ZZFeatureMap to represent input data
fm1 = ZZFeatureMap(2)

# Create the feature map, composed of our two circuits
fm = fm0.compose(fm1)

for trainable_kernel_type in [TrainableFidelityQuantumKernel, TrainableFidelityStatevectorKernel]:

    start_time = default_timer()

    # Instantiate quantum kernel
    quant_kernel = trainable_kernel_type(feature_map=fm, training_parameters=training_params)

    # Set up the optimizer
    spsa_opt = SPSA(maxiter=10, learning_rate=0.05, perturbation=0.05)

    # Instantiate a quantum kernel trainer.
    qkt = QuantumKernelTrainer(
        quantum_kernel=quant_kernel, loss="svc_loss", optimizer=spsa_opt, initial_point=[np.pi / 2]
    )

    # Train the kernel using QKT directly
    qka_results = qkt.fit(X_train, y_train)
    optimized_kernel = qka_results.quantum_kernel
    print(qka_results)

    # Use QSVC for classification
    qsvc = QSVC(quantum_kernel=optimized_kernel)

    # Fit the QSVC
    qsvc.fit(X_train, y_train)

    # Predict the labels
    labels_test = qsvc.predict(X_test)

    # Evaluate the test accuracy
    accuracy_test = metrics.balanced_accuracy_score(y_true=y_test, y_pred=labels_test)
    print(f"Accuracy test: {accuracy_test}")

    # Compute run time
    end_time = default_timer()
    run_time = end_time - start_time
    print(f"Run time for {trainable_kernel_type.__name__}: {run_time} secs")

Output:

{   'optimal_circuit': None,
    'optimal_parameters': {ParameterVectorElement(θ[0]): 1.7210511412417313},
    'optimal_point': array([1.72105114]),
    'optimal_value': 14.249934725144328,
    'optimizer_evals': 20,
    'optimizer_result': None,
    'optimizer_time': None,
    'quantum_kernel': <qiskit_machine_learning.kernels.trainable_fidelity_quantum_kernel.TrainableFidelityQuantumKernel object at 0x1593aa680>}
Accuracy test: 1.0
Run time for TrainableFidelityQuantumKernel: 25.38589504099218 secs
{   'optimal_circuit': None,
    'optimal_parameters': {ParameterVectorElement(θ[0]): 1.721051141237652},
    'optimal_point': array([1.72105114]),
    'optimal_value': 14.249935954397358,
    'optimizer_evals': 20,
    'optimizer_result': None,
    'optimizer_time': None,
    'quantum_kernel': <qiskit_machine_learning.kernels.trainable_fidelity_statevector_kernel.TrainableFidelityStatevectorKernel object at 0x15a9386d0>}
Accuracy test: 1.0
Run time for TrainableFidelityStatevectorKernel: 0.8832503749872558 secs

[coveralls]

Pull Request Test Coverage Report for Build 5144942540

• 35 of 35 (100.0%) changed or added relevant lines in 4 files are covered.
• No unchanged relevant lines lost coverage.
• Overall coverage increased (+0.06%) to 92.915%

---

Totals
Change from base Build 5134410619:	0.06%
Covered Lines:	2649
Relevant Lines:	2851

---

💛 - Coveralls

[ElePT]

Sorry for my late review, I think that Steve already commented on all the key points, so LGTM. Maybe now that we are at it, we could also sneak in the fix for #649 in this PR?

[woodsp-ibm]

LGTM.

[CLAassistant]

All committers have signed the CLA.

[coveralls]

Pull Request Test Coverage Report for Build 5222812947

• 35 of 35 (100.0%) changed or added relevant lines in 4 files are covered.
• No unchanged relevant lines lost coverage.
• Overall coverage increased (+0.06%) to 92.915%

---

Totals
Change from base Build 5222270613:	0.06%
Covered Lines:	2649
Relevant Lines:	2851

---

💛 - Coveralls