Expressibility-Enhancing Strategies for Quantum Neural Networks. (arXiv:2211.12670v1 [quant-ph])

Quantum neural networks (QNNs), represented by parameterized quantum
circuits, can be trained in the paradigm of supervised learning to map input
data to predictions. Much work has focused on theoretically analyzing the
expressive power of QNNs. However, in almost all literature, QNNs' expressive
power is numerically validated using only simple univariate functions. We
surprisingly discover that state-of-the-art QNNs with strong expressive power
can have poor performance in approximating even just a simple sinusoidal
function. To fill the gap, we propose …

arxiv networks neural networks quantum quantum neural networks strategies