Studying the Impact of Quantum-Specific Hyperparameters on Hybrid Quantum-Classical Neural Networks

In current noisy intermediate-scale quantum devices, hybrid quantum-classical neural networks (HQNNs) represent a promising solution that combines the strengths of classical machine learning with quantum computing capabilities. Compared to classical deep neural networks (DNNs), HQNNs present an additional set of hyperparameters, which are specific to quantum circuits. These quantum-specific hyperparameters, such as quantum layers depth, number of qubits, type of entanglement, type of encoding, number of shots, and measurement observables, can be tuned to modify the behavior of the HQNNs and their capabilities to learn the given task. In this paper, we investigate the impact of these variations on different HQNN models for image classification tasks, implemented on both Qiskit and PennyLane frameworks. We aim to uncover intuitive and counter-intuitive learning patterns of HQNN models within granular levels of controlled quantum perturbations, to form a sound basis for their correlation to accuracy. The outcome of our study opens new avenues for designing efficient HQNN algorithms and builds a foundational base for comprehending and identifying tunable hyperparameters of HQNN models that can lead to useful implementation and usage.