A sensor tagging approach for reusing building blocks of knowledge in Learning Classifier Systems

摘要: During the last decade, extraction and reuse of building blocks knowledge for learning process Extended Classifier System (XCS) in Multiplexer (MUX) problem domain have been demonstrate feasible by using Code Fragment (CF) (i.e. a tree-based structure ordinarily used field Genetic Programming (GP)) as representation classifier conditions (the resulting system was called XCSCFC). However, use may lead to bloating increase time complexity when tree grows deep. Therefore, we proposed novel XCS, named Sensory Tag (ST). The XCS with ST input is XCSSTC. experiments method were conducted MUX domain. results indicate that XCSSTC capable reusing problems. current study also discussed about two different aspects knowledge. Specifically, “attribution selection” part “logical relation between attributes” part.