Tree-Based Methods for Fuzzy Rule Extraction
Shuqing Zeng, Nan Zhang, Juyang Weng · 2005
This paper is concerned with the application of a tree-based regression model to extract fuzzy rules from high-dimensional data. We introduce a locally weighted scheme to the identification of Takagi-Sugeno type rules. It is pro-posed to apply the sequential least-squares method to esti-mate the linear model. A hierarchical clustering takes place in the product space of systems inputs and outputs and each path from the root to a leaf corresponds to a fuzzy IF-THEN rule. Only a subset of the rules is considered based on the lo-cality of the input query data. At each hierarchy, a discrimi-nating subspace is derived from the high-dimensional input space for a good generalization capability. Both a synthetic data set as well as a real-world robot navigation problem are considered to illustrate the working and the applicabil-ity of the algorithm.