论文部分内容阅读
目前有关案例推理(CBR)的研究主要集中在案例检索方面,对案例库构造方法的研究则较为少见,而好的案例库,既可以提高案例检索效率,又可以保证较好的检索准确率,鉴于此,针对CBR中的案例库进行研究,引入模糊C均值方法去除原案例库中的冗余案例,从而实现对神经网络,案例推理方法的改进,最后通过对UCI数据进行的仿真实验表明了改进后的案例推理方法无论在案例检索精度还是在案例检索速度上均有所提高。
At present, researches on case-based reasoning (CBR) mainly focus on case retrieval and research on case-based construction methods is rare. A good case base can not only improve case retrieval efficiency but also ensure better retrieval accuracy, In view of this, we study the case base in CBR and introduce the fuzzy C-means method to remove the redundant cases from the original case base so as to improve the neural network and case-based reasoning method. Finally, the simulation experiments on UCI data show that The improved case-based reasoning method has improved both in case retrieval accuracy and in case retrieval speed.