Pulsar Classification: Comparing Quantum Convolutional Neural Networks and Quantum Support Vector Machines

TL;DR

This study compares quantum convolutional neural networks and quantum support vector machines for pulsar classification, highlighting their performance, training efficiency, and suitability in noisy quantum environments, with classical methods serving as benchmarks.

Contribution

It provides a comparative analysis of QCNNs and QSVMs for pulsar classification, including their performance, training time, and robustness to noise in the NISQ era.

Findings

QCNNs train faster than QSVMs.

QSVMs perform better under noisy conditions.

Classical methods serve as effective benchmarks.

Abstract

Well-known quantum machine learning techniques, namely quantum kernel assisted support vector machines (QSVMs) and quantum convolutional neural networks (QCNNs), are applied to the binary classification of pulsars. In this comparitive study it is illustrated with simulations that both quantum methods successfully achieve effective classification of the HTRU-2 data set that connects pulsar class labels to eight separate features. QCNNs outperform the QSVMs with respect to time taken to train and predict, however, if the current NISQ era devices are considered and noise included in the comparison, then QSVMs are preferred. QSVMs also perform better overall compared to QCNNs when performance metrics are used to evaluate both methods. Classical methods are also implemented to serve as benchmark for comparison with the quantum approaches.