论文部分内容阅读
传统的推理方式存在推理效率低和冲突消解问题．提出一种新的解决这一问题的方法．即由正向神经网络推理和反向逻辑推理所组成的混合推理系统．利用种经网络所具有的并行性、可学习性、知识可存储性不但可以解决推理冲突，而且还可以利用并行性提高推理的收敛速度和推理效率．为了说明问题，设计了一个实验系统，经实验取得了较理想的效果，并有实用性．
Traditional reasoning methods have the problems of low efficiency of reasoning and conflict resolution. Put forward a new method of solving this problem. That is, a mixed inference system composed of forward neural network reasoning and reverse logical reasoning. Utilizing the parallelism, learnability and knowledge storability of networked species not only can solve inference conflict, but also can improve the convergence rate of reasoning and reasoning efficiency by using the parallelism. In order to illustrate the problem, an experimental system has been designed and achieved satisfactory results and practicality.