Noboru Wakami

A learning method of fuzzy inference rules by descent method

The authors propose a learning method for fuzzy inference rules by a descent method. From input-output data gathered from specialists, the inference rules expressing the input-output relation of the data are obtained automatically. The membership functions in the antecedent part and the real number in the consequent part of the inference rules are tuned by means of the descent method. The learning speed and the generalization capability of this method are higher than those of a conventional backpropagation type neural network. This method has the capability to express the knowledge acquired from input-output data in the form of fuzzy inference rules. Some numerical examples are described to show these advantages over the conventional neural network. An application of the method to a mobile robot that avoids a moving obstacle and its computer simulation are reported. >

Construction of fuzzy inference rules by NDF and NDFL

Abstract Whereas conventional fuzzy reasoning lacks determining membership functions, a neural network driven fuzzy reasoning (NDF) capable of determining membership functions uniquely by an artificial neural network is formulated. In an NDF algorithm the optimum membership function in the antecedent part of fuzzy inference rules is determined by a neural network, while in the consequent parts an amount of reasoning for each rule is determined by other plural neural networks. On the other hand, we propose a new algorithm that can adjust inference rules to compensate for a change of inference environment. We call this algorithm a neural network driven fuzzy reasoning with learning function (NDFL). NDFL can determine the optimal membership function and obtain the coefficients of linear equations in the consequent parts by the searching function of the pattern search method. In this paper, inference rules for making a pendulum stand up from its lowest suspended point ar3 determined by the NDF algorithm for verifying its effectiveness. The NDFL algorithm is formulated and applied to a simple numerical example to demonstrate its effectiveness.

An application of fuzzy set theory for an electronic video camera image stabilizer

An electronic video camera image stabilizer has been developed which eliminates a substantial part of the image instability caused by the involuntary movement of camera holders. For the discrimination between image movement caused by unstable hand-holding and that of moving objects, fuzzy set theory is applied through the following process: 1) dividing the image taken by the camera into four regions, 2) providing two signals to discriminate between the causes of image instability, 3) evaluating these two discriminating signals after they are transformed into reliability values by membership functions, and 4) tuning the membership functions using a simplex method. This image stabilizer has been incorporated into a new compact video camera, and its substantially improved field performance has been confirmed. >

Segmentation of thermal images using the fuzzy C-means algorithm

A segmentation methodology based on the fuzzy clustering algorithm is developed. The algorithm is utilized to segment a thermal image of occupants in a room taken by a thermoviewer. The purpose of segmentation is to identify the number and the positions of the occupants. Some useful applications can be realized, such as control of air-conditioning systems, security systems, and so on. The approach consists of two stages. The first stage is to distinguish occupants from a background in an image using the fuzzy C-means (FCM) algorithm. The authors have selected a suitable measure for determining the number of clusters and modified it for FCM. The purpose of the second stage is to distinguish each occupant by locating local temperature peaks in the image. A region-growing algorithm is introduced for more accurate segmentation based on the membership value determined by FCM and the number of located peaks. Some experimental results are included that relate to thermal images obtained in a meeting room.<<ETX>>

Formulation of CMAC-fuzzy system

To get the input-output data for the identification of an optimum controller, it is necessary to ascertain the consistent intention underlying the possibly inconsistent actions of the human operator. To acquire the intention of the human operator, the authors propose a cerebellar model arithmetic computer (CMAC) fuzzy system to construct fuzzy inference rules that indicates the human intention. The algorithm to construct the fuzzy inference rules is described. The formulation of the CMAC fuzzy system is explained, and an application of the CMAC fuzzy system is shown by taking an example of a computer simulation of catching a moving object.<<ETX>>

An electronic video camera image stabilizer operated on fuzzy theory

An electronic video camera image stabilizer was developed. It eliminates a substantial part of the image instability caused by the involuntary movement of camera holders. The discrimination between image movement caused by unstable hand-holding and that of moving objects becomes possible by developing the following process: (1) dividing the image taken by the camera into four regions, (2) providing two signals to discriminate between the causes of image instability, (3) evaluating these two discriminating signals after they are transformed into reliability values by membership functions, and (4) tuning the membership functions using a simplex method. This image stabilizer was incorporated into a compact video camera and its substantially improved field performance was confirmed.<<ETX>>

Robust PCA with Intra-Sample Outlier Process Based on Fuzzy Mahalanobis Distances and Noise Clustering

To make principal component analysis (PCA) robust for intra-sample noise, Torre and Black proposed a general analogue outlier process that provides a connection to robust M-estimation. This paper proposes a fuzzy membership approach based on the squared Mahalanobis distances and the noise clustering (NC) by Dave for robustizing PCA to intra-sample outliers

A Proposal of a Fuzzy Connective with Learning Function and Query Networks for Fuzzy Retrieval Systems

A new fuzzy connective and a structure of network constructed by the fuzzy connectives are proposed here to overcome a drawback of conventional fuzzy retrieval systems. This network represents a retrieval query. In the network, the fuzzy connectives have a learning function to adjust its parameters for minimizing the square of errors for retrieval results. The fuzzy retrieval system employing this network is also constructed. Therefore, users can retrieve the most suitable results for users.

Fuzzy intelligent information processing in home appliances

We report an application of fuzzy logic to intelligent information processing. We have automated keyword extraction from a large amount of text information on teletext programs or Internet to provide users with a summary. With a part of this technique, we have also produced television sets which can summarize teletext news programs (in Japanese) into topics and display them for the users.

Fuzzy retrieval system employing fuzzy connectives with learning functions and query networks

A new fuzzy connective and network structure for queries which are constructed by fuzzy connectives are proposed to overcome a drawback of conventional fuzzy retrieval systems. In a conventional fuzzy retrieval system, it is quite difficult for a user to obtain the most suitable results since the user cannot start with making up complete queries. In our retrieval system, if a user gives an estimation of fitting samples in a database which fit the users requests, AND/OR operators in queries which are made up of fuzzy connectives are adjusted to represent the users requests. With the adjusted parameters and a network for the query, this fuzzy retrieval system gives results which better satisfy the users requests. The consistencies of samples are also discussed. Inconsistent samples are defined, and an extracting method for inconsistent samples is proposed. The effectiveness of this proposed fuzzy retrieval system is shown through an experiment.<<ETX>>