Quantum-inspired complex convolutional neural networks

Shangshang Shi

doi:10.1007/s10489-022-03525-0

2022-04-07

DOI：10.1007/s10489-022-03525-0

原文链接

原文求助

摘要

En 中文

Quantum-inspired artificial neural network is an interesting research area, which combines quantum computing and deep learning. Several models of quantum-inspired neuron with real-valued weights have been proposed, and they were mainly used to build the three-layer feedforward neural networks. In this work, we improve the convolutional neural networks (CNNs) by utilizing the quantum-inspired way of data representation and convolutional operation. Specifically, we first improve the quantum-inspired neuron by exploiting the complex-valued weights, which have richer representational capacity and better non-linearity. Moreover, we extend the method implementing the quantum-inspired neurons to perform convolutional operations, and naturally draw the models of quantum-inspired convolutional neural networks (QICNNs) capable of processing high-dimensional data. Here five specific types of QICNNs are proposed, which are different in the way of implementing the convolutional layers and fully connected layers. We establish the detail mathematical framework to implement the QICNNs. The performances of accuracy, convergence and robustness of the five QICNNs against the classical counterpart are tested using the MNIST and CIFAR-10 datasets. The results show that (1) the QICNN can achieve higher classification accuracy (up to 99.65%) than the classical CNN when using the MNIST dataset; (2) the QICNN has faster convergence speed, which means that QICNN can be trained easily than classical CNN when they have a similar number of parameters; (3) the QICNN has better robustness in the case of employing different way of weight initialization or rotating the input data. It is expected that our QICNNs can outperform the classical counterparts in more practical learning tasks.

Keyword:

Quantum-inspired CNNs

Complex CNNs

Quantum-inspired neuron

Classification accuracy

Convergence

Robustness