This paper proposes the use of two dimensional display of pathophysiological state of patients based upon the results of discriminant analysis using multiple laboratory data in order to recognize the state of patients objectively and visually. To this purpose we studied pathophysiological states of patients with gout and patients with diabetes mellitus in contrast with states of healthy subjects and derived several state-planes which gives optimal representations on differences between disease states. It is worth noting that these three groups(gout, diabetes mellitus and healthy subjects)were clustered well into three groups on this sort of plane without disease-specific parameters, i. e., serum uric acid level to gout and serum glucose level to diabetes mellitus.
ERG data measured by Riggs et al. was analysed by use of some multivariate techniques. The results were compared with those derived by their trial and error method based on the psychophysical point of view, especially in the estimation of chromaticity diagram. The multivariate techniques applied are Torgerson-Gower method(i. e. principal coordinate analysis), Shepard-Kruskal method(well known nonmetric multidimensional scaling), and a new method proposed by one of the authors, T. I. The features of the three methods were discussed during the course of analysing the ERG data. The chromaticity diagram predicted in the present study satisfied the psychophysical requirements better than that predicted by Riggs et al. In addition, the present analysis allowed to estimate the noise component in the observed ERG data and clarified their latent structures.
This paper deals with analyzing fuzzy data according to fuzzy groups. Fuzzy data are characterized qualitatively in the form of fuzzy sets and fuzzy groups are defined by several fuzzy sets. In order to quantify such fuzzy data, a fuzzy mean, a fuzzy variance and a fuzzy variance ratio are defined. Using these definitions, we extend Hayashi's Quantification Theory Type II to the Fuzzy Quantification Theory Type II which can handle fuzzy data and fuzzy groups. Since the qualitative data might be fuzzy in nature, this fuzzy theory is useful for analyzing fuzzy data. A simple example is described to illustrate the Fuzzy Quantification Theory Type II.
The transition in dietary life has often been dealt macroscopically in relation to changing factors in the society, culture and various other aspects. However, whatever factor it relates to, it is a consumer's subjective need that determines the direction of his eating habits. In order to research the future tendency in the dietary system, I believe we must approach it with the structural analysis on the level of dietary consciousness that inwardly contributes to the system. Therefore, we conducted a survey of dietary consciousness on the price monitors and tried to elucidate its basic structure. As a result of the factor analysis, 10 obvious factors such as simplification and growing interest in home cookery have been extracted. Moreover, as the outcome of the cluster analysis based on the factor scores, various modes of eating habits have been confirmed and we suppose they are useful in the study of the dietary system in future. As a general trend, it is likely that we will move on to a qualitatively richer diet, as well as a diet suited toeach occasion, time and event of a consumer's life style.
In this article, the researches are reviewed that used purusit and compensatory tracking tasks without controlled element in which the command variable was either step or sinusoidal wave. Following an introductory chapter, Chapter 2 and 3 discuss the responses to step and sinusoidal wave respectively. The title of each section of Chapter 2 is as follows: 2.1 Linearity in human input-output system. 2.2 Adequate form of transfer function of human input-output system. 2.3 Determinating of human transfer function by means of t heoretical curve fitting. 2.4 Estimated curve of sinusoidal response computed from step response. The title of each section of Chapter 3 is as follows: 3.1 Specific character of sinusoidal response. 3.2 Transfer function of human inputoutput system. 3.3 Time variance course and non-linearity in dynamic characteristics of human input-output system. 3.4 Estimated curve of step response computed from sinusoidal response. References consist of 19 foreign and 15 Japanese papers.