A basic cognitive experiment has been conducted to observe physiological characteristics of on-line human congnitive behavior. During the experiment, three different types of data were recorded from the subjects : (i) eye movement data by eye mark recorder (EMR) (ii) ECG, EEG, EDA, etc (iii) verbal reports.
This paper presents the system development for automatic protocol analysis and its applications for the subject's verbal reports during the problem solving process. It was found that automatic protocol analysis system is effective in objective and detailed analysis of the verbal reports from the subjects. And from the analysis thus for conducted, it was concluded that the human error would originate in the method how the subject memorizes during his/her problem solving process.

View full abstract

In this paper, for interval parameter matrices, a method of estimating a domain of the eigenvalues on the complex plane is proposed. First, an interval parameter matrix is similarly transformed into a complex matrix whose elements lie on convex polygons on the complex plane. Second, by applying Gershgorin's theorem to this matrix, a sufficient condition for all the eigenvalues lie in a half-plane divided by a line on the complex plane is given. Then a scaling vector is introduced to this condition and an optimal scaling vector to obtain a sharper estimate is examined. By using this result for various gradients of the line, the domain of the eigenvalues can be estimated.

View full abstract

This paper is concerned with the detection of false data, which is essential in data processing such as analysis of cellular proliferation of head and neck cancer. False data include, outliers and missing values. The characteristic of the method presented here is the use of computer graphics of an expert system in order to detect outliers. The determination whether measurement data are outliers or not is done by comparing the statistics with the expected value based on a chi-square test. Two examples are presented. One is normal data and the other is a case containing abnormal data.
We are going to develop an outlier detection supporting. system with an object-oriented approach. The suporting system presented here have various kinds of useful knowledge.

View full abstract

A learning control scheme using a forgetting factor and a long term memory is proposed, in which outputs refer to velocity coordinates and the next command input is modified by the residual error between the desired output and actual output. Robustness problems of the scheme against initialization errors and fluctuations of dynamics are treated rigorously for a class of linear mechanical systems. First, the uniform boundedness of output trajectories during repetitive learning is proved on the basis of the passivity of the mechanical system. Next, the convergence of output trajectories to an &-neighborhood of the desired output trajectory is proved. The size of the E-neighborhood depends on the magnitude of initialization errors and fluctuations of dynamics.

View full abstract

[in Japanese]

View full abstract