By introducing the principle and characteristics of constructive neural networks (CNN) and pointing out its deficiencies, fuzzy theory is adopted to improve the covering algorithms in this paper. We build "extended area" for each type of samples, eliminate the inference of the outlier, and redefine the threshold of covering algorithms. Furthermore, "sphere neighborhood" (SN) are constructed, the membership functions of test samples are given and all of the test samples are determined accordingly. First of all, the procedure of constructive fuzzy algorithm is given, then the model of constructive fuzzy neural networks (CFNN) is built, finally, CFNN is applied to search for communications signals. Extensive experimental results demonstrate the efficiency and practicability of the proposed algorithm. © Springer-Verlag Berlin Heidelberg 2005.
CITATION STYLE
Wang, L., Tan, Y., & Zhang, L. (2005). Constructive fuzzy neural networks and its application. In Lecture Notes in Computer Science (Vol. 3496, pp. 440–445). Springer Verlag. https://doi.org/10.1007/11427391_70
Mendeley helps you to discover research relevant for your work.