A Knowledge Mining Method for Continuous Data Based on Fuzzy C-Means Clustering and Rough Sets

For continuous data knowledge mining, fuzzy c-means clustering is used for dispersing continuous data to discrete variables and then the discretized data can be reduced by using rough set theory. In order to determine corresponding classes, one usually utilizes the maximum membership degree method. However when membership values are close together, above method can´t reflect the essential characteristics of continuous data. In this paper, a new method, which is based on so-called membership superposition degree instead of maximum membership degree, is proposed. The method is applied to the knowledge mining of steamer axes vibration data. Result shows that the method can reflect the characteristics of continuous data