摘要
文章介绍了基于示例学习算法IBL的概况,并对其加以改进,提出了一个新的算法IBL-Cluster。它主要由概念描述形成算法和概念描述修改算法构成。在此基础上建立了应用在基于事例的推理系统CBR中的基于IBL算法的索引与检索机制。实验表明新算法IBL-Cluster在存储空间及测试正确率方面均有改善。
: The instance-based learning(IBL)algorithms are first introduced in this paper.Then a new algorithm IBL-Cluster is presented on the basis of IBL algorithm and it mainly includes concept description formative algorithm and concept description updating algorithm.Finally,the index and retrieval mechanism based on IBL algorithm in CBR system can be given.The test shows that IBL-Cluster is better than IBL about storage requirement and average accuracy.
出处
《计算机工程与应用》
CSCD
北大核心
2001年第5期67-69,95,共4页
Computer Engineering and Applications
基金
哈尔滨工业大学校基金资助。
关键词
人工智能
IBL算法
CBR系统
索引
检索
机器学习
: Instance-based learning(IBL),case-based reasoning(CBR),concept description formative algorithm,concept description updating algorithm.