PREDICTION PHASE IN ASSOCIATIVE CLASSIFICATION MINING

摘要: Associative classification (AC) is an important data mining approach which effectively integrates association rule and classification. Prediction of test a fundamental step in that impacts the outputted system accuracy. In this paper, we present three new prediction methods (Dominant Class Label, Highest Average Confidence per Class, Full Match Rule) one pruning procedure (Partial Matching) AC. Furthermore, review current Experimental results on large English Arabic text categorisation collections (Reuters, SPA) using proposed other popular algorithms (SVM, KNN, NB, BCAR, MCAR, C4.5, etc.), have been conducted. The bases comparison experiments are accuracy Break-Even-Point (BEP) evaluation measures. reveal our very competitive with reference to BEP if compared known AC approaches such as those 2-PS, ARC-BC BCAR. Moreover, outperform existing traditional decision trees, probabilistic regards Finally, indicate improved classifier.