人工智能神经网络在交叉立方体上的有效映射

Efficient Mapping of ANNs on Crossed Cubes

Malluhi等人在文献犤1犦中介绍了人工智能神经网络(ANNs)在超立方体上的有效映射,交叉立方体是超立方体的一个重要变型,而且具有比超立方体更优越的性质,如果在交叉立方体上实现ANNs的有效映射,会有更好的意义。论文证明了一个N×NMAT(mesh-of-appendixed-trees)可以嵌入包含4N2个节点的交叉立方体中,其中N是最大一层的长度,并且证明这个嵌入是最优的,从而给出了ANNs在交叉立方体上的一个有效映射。

The efficient mapping of ANNs on hypercube is introduced in the articleby Malluhi.The crossed cube is an important variant of the hypercube,and it has several attractive properties compared with the hypercube.This paper proves that a N×N MAT can be embedded into a crossed cube and this embedding is the optimal,then the efficient mapping on crossed cubes is given.