作者: Liqiang Geng , Christine W. Chan
关键词:
摘要: Knowledge acquisition for a case-based reasoning system from domain experts is bottleneck in the development process. With huge amounts of data that have become available, it would be useful to derive automatically representative cases available databases rather than acquiring them experts. This paper presents two algorithms using similarity-based rough set theory databases. The first algorithm SRS1 requires user decide similarity thresholds objects database, while second SRS2 can select proper thresholds. These require fewer parameters other case generation algorithms. Also they tackle noise and inconsistent database reasonable number database. experimental results were compared with those well-known mining systems, such as rule induction systems neural network systems.