<ul><li>Rule representations significantly impact the search capabilities and decision boundaries in Learning Classifier Systems (LCSs).</li><li>Adaptive mechanism introduced to select appropriate rule representation for each rule in LCSs based on different subspaces within the input space.</li><li>Flexible rule representation using a four-parameter beta distribution integrated into a fuzzy-style LCS to automatically choose suitable representations for varying subspaces.</li><li>Experimental results on real-world tasks show that the LCS with the new rule representation achieves higher test accuracy and generates more concise rule sets.</li></ul>

Adapting Rule Representation With Four-Parameter Beta Distribution for Learning Classifier Systems

Discover more