Abstract:

This paper presents the formulation of additive Fuzzy adaptive model by using the framework of Gaussian Mixture Model, which provides the membership functions for the input fuzzy sets. The consequent part of the model is the output function which is derived from the adaptable parameter vector consisting of a weight of a rule, mean and covariance as its elements. These elements are updated using the Expectation and Maximization (EM) algorithm which is equivalent to Baum- Welch’s backward and forward algorithm for estimating Hidden Markov Model parameters. This resulting model is found to be adaptable depending on the desired input-output behavior. The model has also been tested on a benchmark problem and the results1 are found to be better than those obtained from the well known fuzzy models including additive fuzzy models..

Keywords: GMM, Additive fuzzy systems, EM and GFM.

(Full Paper 413KB)

BibTex:

@ARTICLE{P1120615103,

AUTHOR = {Nishchal K. Verma, M. Hanmandlu},

TITLE = {Adaptability in Additive Fuzzy System via EM Algorithm},

JOURNAL ={The International Journal of Artificial Intelligence and Machine Learning},

}

(Full Paper 413KB)