AAAA AAAA AAAA AAAA Coil - Car AAA AAA AAAA AAAA A A A AAA
چکیده
An accurate retrieval system is essential to the performance of any case-based reasoning (CBR) system. This paper describes a new retrieval method in which both feature similarity and adaptation knowledge are used to guide retrieval. A system can only improve its performance by learning and CBR systems learn mainly by storing newly adapted cases. The system has better opportunities to learn useful cases if retrieval is accurate. Thus retrieval is a foundation for learning in CBR systems.