Modification of Membership Function in Fuzzy Least Square (FLS) Method and Its Merge with Sugeno Fuzzy Inference System for Data Uncertainty Process

الملخص

Data is often encounters problems during collection and tabulation as existence of missing, outlier, or uncertainty values. This would affect the validation of the results.

This paper provides uncertainty study and its process using proposed method. This method for uncertainty values corrections is proposed using both Sugeno fuzzy inference system and (FLS) fuzzy least square. As soon as modify of membership function in fuzzy least square, with fuzziness of dependent variable data, that work to minimize spread values in the target function to obtain the coefficients which approximate the uncertainty values to their original values.

In this paper, the proposed idea is applied to a sinusoidal signal using MATLAB® toolbox with additive uncertainty. The application of the proposed method shows that the accuracy of the correction is 99% using two statistical criteria; the mean square error (MSE) and mean absolute percentage error (MAPE) compared to the traditional least square method used by the Sugeno inference system.