Fuzzy Rule Generation based on coWM and FCM Algorithms
Feng Hou, Jin Gou · International Journal of Applied Mathematics & Statistics/International journal of applied mathematics and statistics · 2013
The sample in fuzzy rule generation which contains noise and outlier may result in invalid rules and the confidence of the rules generated by the WM algorithm could be low. To address these problems, a new method for fuzzy rule generation is proposed in this paper. In the first stage, the Fuzzy C-Means clustering (FCM) algorithm is used to optimize the original sample. Then, only the optimization sample set is exploited to generate the rules by utilizing the proposed coWM method, which is improved based on the WM method by considering the cooperation among input variables. Experimental results have shown that the proposed method is not only robust to the noise and the outlier contained in the sample but also achieves the better approximation performance of the fuzzy system. Furthermore, it should be pointed out that the proposed method is able to reduce the computational complexity of the computational model, since the amount of the input optimized sample is smaller than that of the original samples.