利用分治技术的智能推理

Artificial Reasoning Using the Divide-and-Conquer Technique

一、引言“分治”技术是一种分解-合成技术,即对于复杂的不能直接求解的问题,将该问题分割成几个子问题,分别求解子问题,然后合成子问题的解,以获取整个问题的解。若分割出的子问题仍不易求解,则继续分解子问题,一直到可以直接求解为止。因此,这是一个递归模式。同样,人工智能在求解难题时,常常将难题分解为较为简单的子问题,然后合成各个子问题的解决方案,形成一个整体方案。本文通过描述分治技术在基于类比推理、基于事例推理以及基于原型推理的应用,表明了该技术在智能推理中的有效性和灵活性。

This paper introduces the divide-and-conquer technique for the artificial reasoning. The effectiveness and the flexibility of the divide-and-conquer technology are demonstrated in the field of the artificial reasoning by showing the technology applications in analogy-based reasoning,case-based reasoning and prototype-based reasoning.