ABSTRACT
This chapter discusses an algorithm to generate a fuzzy learning decision tree by determining its structure and parameters. This algorithm first collects enough training data for generating a practical decision tree. The chapter utilizes fuzzy statistics to calculate fuzzy sets for representing the training data in order to save computation memory and increase generation speed. It also utilizes a suboptimal criterion to determine a decision tree from the resultant fuzzy sets. The chapter introduces tactile sensors and the construction of a tactile sensing, and recognition system. It describes tactile system that classifies geometric objects and human hand using the proposed fuzzy learning decision tree. Decision trees can be applied to many fields such as pattern recognition, classification, decision support systems, expert systems, robot control, and dynamics identification. In some prediction and recognition applications, decision trees have comparable performance with artificial neural networks.
