This paper proposes a new technology, “a bodygraphic information system (BIS)”, that represents human-body-related information based on a human body coordinate system. This representation allows us to accumulate human-body-related information in a standardized and multilayered way. Standardized and multilayered representation enables retrieval, share, statistical analysis, and integration of human-body-related information in different fields such as medicine, engineering, and industry. This paper describes the prototype of BIS and also presents the feasibility and the significance of BIS by applying it to childhood injury surveillance in cooperation with a hospital. Using the developed BIS, the authors collected 1,628 children's injury data and visualized statistical analysis of the data.
As an experimental study for balancing daily human comfort and energy saving, this paper describes a structural equation modeling for questionnaire data. The goal is to clarify the causal relation among thermal comfort and thermal sensation of body parts. The experimentation data include 22 experimental subjects' votes for their feeling on nine conditions. There are four alternatives for variables in modeling: (1) thermal sensation vote (TSV) of time t, (2) TSV of time t and time t-10, (3) TSV of time t and its change, (4) change of TSV. Then numerical study got three findings as follows: (1) The fourth model dealing with the changes of thermal sensations has the fittest indices, (2) Change of thermal sensations of buttock and chest have great impact on comfort, and (3) Change of fatigue impression has no relation with thermal sensations.
Present methods of service provision have proven insufficient to guide users efficiently to the mobile internet services they need. To solve this problem, the authors have been investigating a task-oriented menu which enables users to search for services by “what they want to do” instead of by “name of category”. Construction of such a task-oriented menu is based on a task ontology modeling method which supports the description of user activity such as task execution and the solving of obstacles encountered during the task. We also discuss a modeling method which supports the description of users' activity and related knowledge.
Understanding everyday life behavior requires multidisciplinary and integrated research. This stems from the fact that everyday life behavior is much higher dimensional phenomena than a single academic society can deal with. This paper discusses a method for representing and processing information to deal with everyday life behavior and proposes a new concept of a “spatio-temporal-semantic mapping system (STS Map)” that allows us to standardize and integrate a variety of information and to share information multidisciplinary. The features of the STS Map lie in 1) spatio-temporal expansion of behavior metrics sensing with location sensor, 2) representation based on an environmental coordinate system, 3) a statistical modeling process for knowledge acquisition and a retargeting process. To show the effectiveness of the proposed STS mapping system, this paper reported an implemented system and a case study using the implemented system. In the case study, we conducted in situ measurement of 47 children playing with equipment by a wireless wearable location-electromyography sensor and created two models on a children's skill from the measured data. This paper reported cross-validation of the constructed behavior models and compared them. This paper also reported a new design of play equipment having a climbing part that was suitable for the target age group of children. The new design was conducted using the created model in collaboration with playground equipment designers.
In this paper, we propose an automated marketing information system for recognizing customer buying behavior in a store using ultrasound sensor and support vector machine. Customer buying behavior means the acts of person involved in buying decision processes at selling area of retailers such as supermarket and shopping center. For example, customer stops in front of shelf, and stretches hand to grasp goods etc. These behaviors express signals of customers' interest in goods. In Japan, many retailers are in face of a difficult problem of low productivity of selling area. One of the biggest problems is that most retailers tend to overlook their customer's behaviors. The retailers must be aware of these customer behaviors as important marketing data in order to make the optimum sales promotion that is well suited to their customer behavior. By some experimental results, we verified that the proposed system can recognize customer buying behavior. The use of our marketing information system will do a lot for making a good example of raising the productivity of selling area using information technology.
This paper, by paying attention to human purchasing behavior, attempts to establish a fuzzy reasoning model so that researching the marketing activities of enterprises and demand forecasts will be smoothly carried out. Concretely, the writers select meteorological factors, which would be useful by applying the principal component analysis, and also try to analyze the correlation between the number of visitors and the meteorological factors. Then, by applying an analysis of variance, they attempt to analyze day-to-day sales and the relationship between the number of visitors and the meteorological factors. On the basis of the result obtained from this fuzzy reasoning model, in which demand forecasting can be done, is constructed by selecting the meteorological factors which influence purchasing activities. Finally, the writers try to set up a presupposition for demand forecast. At the same time, they compare the forecast results of a regression analysis that is an earlier forecasting technique in order to clarify the utility of their fuzzy reasoning model.
Recently fuzzy association rule mining techniques have been applied to classification problems to design an accurate and compact classifier. A fuzzy association rule is a way of describing a relationship in the form of if-then statements with fuzzy sets. To define the interestingness of fuzzy association rules, several criteria such as confidence and support have often been used. Association rule mining aims to extract all rules that satisfy pre-specified thresholds of confidence and support. When building a classifier from extracted fuzzy rules, it is impossible to examine all of their possible combinations. Genetic fuzzy rule selection deals with such a problem using genetic algorithms. It efficiently searches for an accurate and compact classifier. In this paper, first we demonstrate the performance of genetic fuzzy rule selection. Next, we examine obtained classifiers by genetic fuzzy rule selection. Experimental results show that Pareto-optimal rules in terms of confidence and support are often included in the obtained classifiers. Based on these observations, we propose and examine genetic fuzzy rule selection that selects fuzzy rules from only Pareto-optimal rules. Through computational experiments, we show that genetic fuzzy rule selection with only Pareto-optimal candidate rules often has almost the same accuracy and much less computation time than the case with all candidate rules.
A fuzzy trend model is introduced for analyzing the trend of long-term financial time series based on the Takagi-Sugeno's fuzzy system. Most of financial time series models are focused on the variance of time series under the assumption that the mean is constant or has a special structure. However, such an assumption does not hold for long-term time series. On the other hand the mean value of the fuzzy trend model can fluctuate slowly. In this paper an identification plocedure of the fuzzy trend model is proposed and its usability is examined by simulation studies. Moreover the proposed plocedure is applied to Tokyo Price Index (TOPIX) and it is shown that the fuzzy trend model provides a new correlation analysis method on the trend.
This paper proposes a simultaneous application of homogeneity analysis and fuzzy clustering which simultaneously partitions individuals and items in categorical multivariate data sets with possibilistic constraints. In the proposed method, the graded possibilistic approach is applied to estimation of memberships of items for deriving the absolute responsibilities of them. The memberships can be regarded as the probability that an experimental outcome coincides with one of mutually independent events. Then, soft transition of memberships from probabilistic to possibilistic constraint is performed by using the graded possibilistic constraint in the approach. We demonstrate the feasibility of our method in the application to the questionnaire data.
This article presents a new approach of the morphological associative memory (MAM) without a kernel image to improve the perfect recall rate by introducing the small world network. The MAM is one of the powerful associative memories compared to ordinary associative memories in terms of calculation amount, memory capacity and noise tolerance. However, the MAM needs the kernel image which is susceptibility to noise and difficult to design. Although, as a practical model, the MAM without a kernel image has been proposed, the model has a problem that the perfect recall rate is degraded. On the other hand, it has been reported that an introduction of the small world network to associative memories is effective in the recall rate improvement and the network size reduction. The small world network is easy to handle because it can design by β-Graph which is controlled by only one parameter. We try to improve of the perfect recall rate and reduce the network size by using the small world network. The effectiveness of the proposed approach is confirmed by autoassociation experiments.
In this paper, we propose a stochastic binary search algorithm utilizing weighted averaged evaluation value calculated from search history. By applying the proposed method to adjustment of the positioning and angle of the optical axes, we quantitatively verify that the proposed method can drastically reduce adjustment time and improve the precision of adjustment. Adjustment time can be reduced by the stochastic binary search algorithm and the precision of adjustment can be improved by adopting the weighted averaged evaluation value calculated from search history. Adjustment experiments with the proposed method demonstrate that the proposed method was able to reduce the adjustment time from 3 hours to 12 minutes and to adjust the positioning and angle of the optical axes robustly in noisy environments.