论文部分内容阅读
定义人工神经网络为具有二重属性的一阶对象。当用于活动模型时作为函数；当着重内部状态时视为结构。这种定义除了其良好的结构之外，还具有重要的优点，便于神经网络的合成；易于集成进行推理。本项研究的一个主要目的是开发与传统符号推理相适应的定义方法，以及应用神经网络的环境
Define artificial neural network as a first order object with double property. Used as a function when used in an active model; considered as a structure when focused on internal states. This definition, in addition to its good structure, has important advantages for the synthesis of neural networks; it is easy to integrate for reasoning. One of the main purposes of this study is to develop a definition method that is compatible with traditional symbolic reasoning and the environment in which neural networks are applied