Minimal subset generation on Lazy Associative Classification based on cogency and Harmonic Mean

Document Type : Primary Research paper


1 Assistant Professor, Department of computing science and Engineering, Kumaraguru College of Technology, Coimbatore, India

2 Professor, School of Computer Science and Engineering, Vellore Institute of Technology, Chennai Campus, India


Lazy learning associative classification method gains higher accuracy than counterpart eager method. The majority of the existing lazy associative classification algorithms generate an exponential number of subsets that increase the computation time. Moreover, lazy learning associative classification with traditional support and confidence measure leads to missed out some rare and prime subsets. The proposed method overcome this problem by focusing on the important feature of the given test instance. As a outcome the proposed system is able to generate a minimal number of high-quality rare itemset as a subset. The proposed rare lazy associative classifier produces high quality subsets and increases classifier accuracy, according to the empirical results. Experiment results show that the proposed algorithm attains better accuracy and reduce the computation time than traditional lazy learning associative classification.