In this paper, a fuzzy classification association rule (FCAR) model is proposed based on the improved G-K algorithm, improved multiple supports and the proposed fuzzy associative classification rules (FACR) approaches to improve the prediction accuracy. The improved G-K algorithm is used to define the membership functions of fuzzy sets, while FACR improves current associative classification approaches by adapting the improved multiple support algorithm. The proposed FCAR model can provide a generalized prediction model to deal with different application domains. The validation of the FCAR model is conducted using different benchmark data sets from the University of California, Irvine (UCI) of machine learning and knowledge extraction based on evolutionary learning (KEEL) repositories, then the results of FCAR are compared with common prediction model artificial neural network (ANN). The experimental results show that the proposed model discovers rules that effectively minimize the prediction error rate.
|Title of host publication||Software, Knowledge, Information Management and Applications (SKIMA 2013)|
|Subtitle of host publication||Advanced Technology Solutions and Applications in Higher Education and Enterprises|
|Editors||Abdelaziz Bouras, Pitipong Yodmongkon, Keshav Dahal, Yacine Ouzrout, Napaporn Reeveerakul|
|Publisher||Inderscience Enterprises Limited|
|Number of pages||11|
|Publication status||Published - 2013|
Sowan, B., Dahal, K., & Hossain, A. (2013). A prediction model based on multiple support and associative classification approaches. In A. Bouras, P. Yodmongkon, K. Dahal, Y. Ouzrout, & N. Reeveerakul (Eds.), Software, Knowledge, Information Management and Applications (SKIMA 2013): Advanced Technology Solutions and Applications in Higher Education and Enterprises (pp. 292-302). Inderscience Enterprises Limited.