Extracting IF-THEN Rules from Numerical Data using Wang-Mendel Methods
Minshen Hao, Jerry M. Mendel · 2013
Abstract Knowledge and experiences are extremely important in the petroleum field. Most of the knowledge is in some expert's mind and it is not easy to validate and use it automatically. In this paper, we introduce an approach to extract linguistic IF-THEN rules from numerical data. These kinds of rules are not only easy to implement in intelligent systems, but are also human-understandable. A soft clustering algorithm, the fuzzy c-mean (FCM) algorithm, is used to determine the MFs of the linguistic terms for each variable. Then, the Wang-Mendel (WM) method and two of its variants are applied to some historical data to extract IF-THEN rules. Finally, the generated rules are compared and validated using the given data.