Skip to Main Content
Decision tree is an effective way of classification rule extraction. For the incomplete decision system, an algorithm of constructing decision tree and classification rule extraction is presented based on the logical relationship between attributes which is described with generalized decision function, and conditional information entropy as heuristic. In the end, experimental results validate its correctness and efficiency by using star spectra data as the incomplete decision system.
Date of Conference: 26-28 Sept. 2010