论文部分内容阅读
距离测度是案例检索的关键问题,它直接影响案例检索精度.针对距离测度进行研究,提出一种基于微粒群方法的自学习距离测度,并将该自学习距离测度引入案例推理中,使案例推理在处理由相关属性表述的案例时有了合理的解决方法,从而扩展了案例推理的应用范围.最后,利用实际数据与UCI数据对基于新距离测度的案例推理技术进行了仿真实验,实验结果表明,与其他方法相比,该方法可以提高案例检索的准确性.
Distance measurement is the key problem of case retrieval, which directly affects the accuracy of case retrieval.Aiming at the distance measurement, a self-learning distance measure based on particle swarm optimization is proposed, and the self-learning distance measure is introduced into case reasoning so that case reasoning In the process of dealing with cases expressed by related attributes, there is a reasonable solution to extend the application of case-based reasoning.Finally, the case-based reasoning techniques based on the new distance measure are simulated by real data and UCI data. The experimental results show that Compared with other methods, this method can improve the accuracy of case retrieval.