作者: Amira Abdel-Aziz , Eyke Hüllermeier
DOI: 10.1007/978-3-319-24586-7_1
关键词:
摘要: In preference-based CBR (Pref-CBR), problem solving experience is represented in the form of contextualized preferences, namely, preferences between candidate solutions context a target to be solved. Since potentially large number such can collected course each episode, case base maintenance clearly becomes an issue Pref-CBR. this paper, we therefore extend our Pref-CBR framework by another component, method for dynamic maintenance. The main goal increase efficiency case-based solving, reducing size base, while maintaining performance. To illustrate effectiveness approach, present study which used repetitive traveling salesman problem.