Complex-valued neural networks, which are extensions of ordinary neural networks, have been studied as interesting models by many researchers. Especially, complex-valued Hopfield neural networks (CHNNs) have been used to process multilevel data, such as gray-scale images. CHNNs with Hermitian connection weights always converge using asynchronous update. The noise tolerance of CHNNs deteriorates extremely as the resolution increases. Noise tolerance is one of the most controversial problems for CHNNs. It is known that rotational invariance reduces noise tolerance. In this brief, we propose symmetric CHNNs (SCHNNs), which have symmetric connection weights. We define their energy function and prove that the SCHNNs always converge. In addition, we show that the SCHNNs improve noise tolerance through computer simulations and explain this improvement from the standpoint of rotational invariance.
In order to investigate the electronic properties of the semiconducting van der Waals ferromagnet (CGT), where ferromagnetic layers are bonded through van der Waals forces, we have performed angle-resolved photoemission spectroscopy measurements and density-functional theory calculations. The valence-band maximum at the point is located below the Fermi level, consistent with the semiconducting property of CGT. Comparison of the experimental density of states with the DFT calculation has suggested that Coulomb interaction between the Cr electrons . The calculation indicates that magnetic coupling between Cr atoms within the layer is ferromagnetic if Coulomb is smaller than 3.0 eV and that the interlayer coupling is ferromagnetic below . We therefore conclude that, for deduced by the experiment, the intralayer Cr-Cr coupling is ferromagnetic and the interlayer coupling is near the boundary between ferromagnetic and antiferromagnetic, which means experimentally deduced is consistent with the theoretical ferromagnetic condition.
A neural network is referred to as minimal if it cannot reduce the number of hidden neurons that maintain the input-output map. The condition in which the number of hidden neurons can be reduced is referred to as reducibility. Real-valued neural networks have only three simple types of reducibility. It can be naturally extended to complex-valued neural networks without bias terms of hidden neurons. However, general complex-valued neural networks have another type of reducibility, referred to herein as exceptional reducibility. In this paper, another type of reducibility is presented, and a method by which to minimize complex-valued neural networks is proposed.