Fuzzy SOFM and its application to the classification of plant communities in the taihang mountains of China

Abstract

Fuzzy set theory and neural network are both attractive for ecological investigations for their powerful in analyzing and solving complicated and non-linear matters and for their freedom from restrictive assumptions. The combination of fuzzy set theory and neural network may produce better technique. This study tried to combine them in clustering analysis of plant communities in the Taihang Mountains in China. The dataset was consisted of importance values of 88 species in 68 samples of 10 m × 20 m. First, we calculated fuzzy similarity matrix of samples; second, the fuzzy similarity matrix was input to neural network; and then the self-organizing feature map (SOFM) was used to classified samples. We called this technique Fuzzy Self-organizing Feature Map (F-SOFM). The 68 samples were clustered into 8 groups, representing 8 vegetation formations. This classification result was reasonable and fully interpreted, which suggests that F-SOFM is effective method in ecological study. The F-SOFM shares both advantages of fuzzy set theory and neural network, and is applicable to all branches of science. © Springer Science+Business Media B.V. 2008.