Kazuya Kishida

A destructive learning method of fuzzy inference rules

In order to construct a fuzzy system with a learning function, numerous studies combining fuzzy systems and neural networks (or descent method) are being carried out. The self-tuning method using the descent method has been proposed by Ichihashi et al. (1991) and it is known that the constructive method is more powerful than other methods using neural networks (or descent method). But this method does not have a sufficient generalization capability or an expressing capability for the acquired knowledge. In this paper, we propose a new learning method called a destructive method of fuzzy inference rules by the descent method. And we show that the destructive method is superior in the number of rules and inference errors but inferior in learning speed to the constructive one. Further more, in order to improve learning speed, we propose a learning method combining the constructive and the destructive methods. Some numerical examples are given to show the validity of the proposed methods, and applications of these methods to the obstacle avoidance problem are shown.<<ETX>>

A Learning Method of Fuzzy Inference Rules Using Vector Quantization

Some models using self-organization systems of neural networks are proposed in recent studies. These models show good results in point of the number of fuzzy rules in high dimensional problems. However, most of these models determine a distribution of initial fuzzy rules by considering only input data. In this paper, we propose a method considering not only input data but also output data. In order to demonstrate the validity of the proposed method, some numerical examples are performed.

A self-tuning method of fuzzy modeling with learning vector quantization

We propose a self-creating method of fuzzy modeling with learning vector quantization. A self-creating neural network is used for vector quantization. There are many fuzzy models using self-organization and vector quantization. It is well known that these models effectively construct fuzzy inference rules representing distribution of input data, and are not affected by increment of input dimensions. We use a self-creating neural network for constructing fuzzy inference rules. In order to show the validity of the proposed method, we perform some numerical examples.