论文部分内容阅读
逻辑守恒性泛指广义的不变性和守恒性。泛权网场复合中的逻辑守恒性则包括稳定集及输入输出关系等。泛权网场进行复合的概念本身也是一般的网络、图、有限自动机、诊断模型等中的运算、状态转移等概念的概括和推广。在这篇文章中 ,首先引进并定义了泛权网场复合的一些规则 ,然后证明了一组与这种抽象的泛权网场复合相关的逻辑守恒性定理。并结合人工神经网络中的均场模型 ,探讨了这种泛权网场在ANN中的描述力及应用前景
Logical conservation generally refers to the generalized invariance and conservation. The logic conservation in the plenoptic network consists of stable set and input-output relations. The concept of compounding in the field of plenoptic network is also the generalization and popularization of concepts such as operations in general networks, graphs, finite automata, diagnostic models, and state transitions. In this paper, we first introduce and define some rules for the composition of plenoptic networks, and then prove a set of logical conservation theorems that are complex with this abstract plenoptic network. Combined with the mean-field model in artificial neural network, this paper discusses the descriptive power and application prospect of this generic network in ANN