In recent years, a high-speed acceleration simulation with a road traffic simulator is needed because of a large-scale traffic jam calculation etc. To satisfy the demand of this society, we improve microscopic road traffic simulator MITRAM, and have developed parallel MITRAM that can simulate a larger-scale region. However, an existing parallel system hardly corresponded to a complex road connection, and it was difficult to simulate it in a complex urban area. In the present study, it aims developing a new connection method for this parallel MITRAM, and doing with parallel MITRAM even what complex and vast urban area simulation it is. Moreover, the performance comparative simulation is done in parallel MITRAM and usually, and how the improvement of the acceleration performance and the load reduction is verified.

View full abstract

When the robotic manipulator performs by voice instructions of a user, interpretations of these instructions include fuzziness because of various qualitative words. The fuzziness in interpretations is an important challenge in the voice control system. One of voice control systems is a fuzzy coach-player system which is quantified ambiguous commands by fuzzy reasonings. In this paper, we propose an experiential update algorithm of fuzzy rules in the fuzzy coach-player system. The effectiveness of the present algorithm is discussed through some experimental results of a robotic forceps.

View full abstract

Assembly mechanisms of mobile robots are a wheel, a crawler or a leg. The legged robot has various advantages to move on an irregular terrain. The central pattern generator (CPG) is a one of method to make the locomotion of a legged robot. Controlled parameters are amplitudes and frequencies of output signals. These parameters depend on the structure of CPG, connection weights and time constants. In this paper, we consider with a multi-legged robot which has a jumpable mechanism. The periodic jumping movements are generated by CPG composed of Matsuoka oscillators. The designed CPG is verified by some simulations.

View full abstract

In the fields of security and robotics, an omni-directional camera, which can take a whole angle (360 [deg]) image, and a PTZ (Pan-Tilt-Zoom) camera, which can move the camera head in horizontal and vertical planes, are employed for visual systems. The omni-directional camera can take images in all direction in one cycle. On the other hand, the PTZ camera can visualize distant objects using the zoom function. However, its dead angle is also larger than that of the omni-directional camera. Here, we propose an OPTZ system (Omni-directional Pan-Tilt-Zoom camera), which combines a hyperbolic mirror used for omni-directional camera and one PTZ camera. OPTZ consists of both a hyperbolic mirror and PTZ camera that can take the image of the mirror. OPTZ realizes the function of both taking an overview image and focusing on a particular distant object. We performed an experiment to detect the ball to confirm the performance of the OPTZ and detection system proposed above. In this report, the images were obtained from the OPTZ mounted on the soccer robot.

View full abstract

In this report, a new noise removal method in color images is presented. In the conventional noise removal methods for color images, the monochrome image processing techniques have been often performed for each color component separately. However, the component-wise noise removal does not give desirable results because the output values may be with color distortion. Therefore, it is desirable to be processed by the vector-type filters in general. In the proposed method, firstly, a tentative output image, in which the noise is removed almost perfectly, is obtained by using the iterative robust switching vector median-based vector $\varepsilon$-filter. Then the residual components between the input and the tentative output images are calculated, and image components consisting of edges, corner and other image details are extracted from the residual components by using the correlation characteristic in RGB components. Consequently, a final output is obtained by adding the extracted image components into the tentative output image. The effectiveness of the proposed method is illustrated through some experiments.

View full abstract

Mobile devices have a function of character recognition in recent years. Due to the function to acquire text information, it is possible to use the mobile devices as the way to use geographical information system, image search and classification engines, etc. However, it is difficult to recognize characters in natural scenes, because there is too much information in these scenes. In the earlier work, a binarization-based method to extract a signboard from those scenes has been reported. There is a disadvantage that this technique depends on colors of text and background. In this study, we present a method of signboard extraction that doesn't depend on the color of text and background. The proposed method employs a stereo image to extract a signboard, and performs a three-dimensional analysis of the scene. The effectiveness of the proposed method was shown by applying to some natural scenes.

View full abstract

There are a lot of people who feel the inconvenience in daily life because of their visual disturbance that is called "low vision", such as old sight and cataract. Compensating the differences of visual perception, e.g. blurring, visual field limitation etc., a method for extracting and presenting visual information is investigated in this report.

View full abstract

A practical contour detection for a moving object using a half-cosine wavelet network and an active contour model (Snakes) is proposed. Generally, on the Snakes, an automatic setting of the initial contour line is not easy. On the other hand, we developed a practical moving object's area, which can describe the area with basis functions. In the proposed method, the initial contour line can be automatically defined by using the center coordinates of the basis functions. The validity of the proposed method is confirmed by contour detection experiments for a moving object in video frames.

View full abstract

In this paper, we get an inferense solutions for natural language propositions involving fuzzy quantifiers and truth qualifiers by using genetic algorithms. We apply shift type transformation for fuzzy predicate. We find the inference solutions are independent on both gene number N and high level gene number h. We use both straight line function predicate and S-function predicate. We get shift type inference solution for straight line function predicate. We find that shift quantity is larger for the inference solution of S-function predicate than that of straight line predicate.

View full abstract

Recently, the reduction of distribution cost is needed to meet both lower prices and higher quality of the product. The expansions of the Internet mail order system make a demand on warehouses to handle many kinds and small quantity of commodities. The picking operations in the warehouses are therefore depending on humans to be responded flexibly. Main activity of the shipment is picking. Walk time accounts about 52% of the total shipping time, so that selecting an appropriate route in the order-picking contributes the decrease of the work cost. In this study, we construct a working model of order pickers in the warehouses handling for the many kinds and a small quantity of commodities, and we apply the GA to an automatic routing method in order to reduce order-picking tour's timescales. Our experimental results suggest that the proposed method provides results suitable for selecting picking course.

View full abstract

Recently, Genetic Algorithm has been studied as an effective approach for large scale optimization problems. However, we have issues of early convergence and settings of many parameters in GA approach. In order to deal with such issues, parameter free genetic algorithm(PfGA) and distributed genetic algorithm(DGA) are proposed. In this paper, we propose distributed parameter free genetic algorithm(DPfGA) that keeps parameter free characteristic and improve efficiency of optimization. We show effectiveness of the algorithm through application of the algorithm to TSP(Traveling Salesman Problem).

View full abstract

In our parallel distributed genetic fuzzy rule selection, a training data set and a population are divided into subgroups. The computation time can be faster by assigning a pair of a training data subset and a sub-population to each of multiple processor cores. To avoid the overfitting of the sub-population to the corresponding training data subsets, the rotation of training data subsets is applied periodically. In this paper, we examine the effects of the rotation operation on the evolution process of the sub-populations and the generalization ability by monitoring the change in the fitness of each sub-population.

View full abstract

A method for removing unnecessary obstacles from urban scenes which are taken by on-vehicle video camera is proposed. Such unnecessary obstacles include: electric wires, telephone poles, roadside trees, pedestrians, and other foreground objects. The removal of these foreground objects can be achieved by using epipolar plane image analysis (EPI). Although the EPI has originally been proposed to remove the foreground objects, it is unsuitable to apply as is to the urban scenes whose background depths change according to the location of the background building. In this paper, we improve the EPI so that multiple background depth can be treated automatically and present a method for removing unnecessary obstacles and interpolating with consistency. Experimental results showed the superiority in comparison with the conventional EPI-based method when we applied the both methods to a simulated urban scene taken in a miniature urban model.

View full abstract

Cortical dysplasia is a congenital anomaly of cerebral cortex. In pediatric, about 80% intractable epilepsy patients are accompanied with cortical dysplasia. Therefore it is very important to detect cortical dysplasia lesions in pediatric brain MR images. However, image features on the pediatric brain MR images of cortical dysplasia aren't clarified. This paper proposes some image features to estimate cortical dysplasia.

View full abstract

ACL (Anterior Cruciate Ligament) reconstruction is performed to recover the injured knee which often happens in sports activity such as skiing, and basketball. The ACL reconstruction makes one or two bone tunnel(s) in the femur and tibia. Then, the harvested graft passes the bone tunnels so that the femur and tibia are connected. The graft is fixed to the femur by using EndoButton. However, loosening of the fixation sometimes happens due to transposition of the EndoButton. It results in instability of the knee motion. Occurrence of the fixation loosening affects clinical outcome. Therefore, it is important to analyse an EndoButton postion with high accuracy. This study set the goal to propose an analysis method of the EndoButton position with registration in MDCT images. In the experiments, the proposed method was applied to 5 (single bundle) and 8 (double bundle) sets of MDCT images. Each set consists of two different direction posing of the same patient in one day. EndoButton of AM shows a distinctive time-depend movement. The flexion movements of single and AM increase with time. As a result, the proposed method could calculate the EndoButton positions in MDCT images based on the unified bone coordinate system which is made by registration.

View full abstract

It is effective to evaluate magnetic resonance (MR) images for diagnosing neonatal cerebral disorders because they often accompany the deformation of the brain shape. However, there are many difficulties when radiologists manually extract cerebral surface from the MR images because of many sectional images over one hundred images and intricate sulci. In addition, there is inter- and intra- reader variability of image interpretation performance. Therefore, it requires automated extraction of cerebral surface from neonatal MR images. There are many methods to extract cerebral surface from adult MR images, however there are few methods for neonatal MR images. This paper proposes a new extraction method based on particle method. The proposed method introduces three kinds of particles corresponding to cerebrospinal fluid, gray matter and white matter. First, particles are assigned according to the cerebral shape. Second, particles are moved to form the homogeneous particles, and are transited to the other particles with respect to MR signal and particle distribution. The proposed method was applied to neonatal MR images. The results showed that the proposed method extracted cerebral surface with high accuracy.

View full abstract

Recently, Japan has become the best country which prolongs life expectancy. It results in increasing of the number of Total Knee Arthroplasty (TKA). In the TKA, appropriate distribution of a force given by a soft-tissue around the knee into the medial and lateral component of the tibial plateau affects good outcome after the surgery. This paper examined to develop the evaluation system which provides a quantitative criterion for soft-tissue balance. The trial insert is used for measuring the appropriate prosthesis size. After measuring the size, a polyethylene-surface, that is one component of the total knee prosthesis, is fixed to a tibial component. The experiment was performed at the valgus with -10, -5, 0, 5, and 10 degrees. In all the experiments, the loads were 0.0 kg, 4.0 kg, 8.0 kg, 12.0 kg, 16.0 kg, and 20.0 kg. In conclusion, the experiments indicate that the proposed system could acquire a force distribution of the measurement trial insert quantitatively, and visualize the soft-tissue balance. Future works are to improve the resolution and accuracy, and to apply to more real situation.

View full abstract

Color deficiencies have a specific color vision that some colors are indiscriminable for them. Therefore, they sometimes cannot receive the information expressed by differences of colors. To cope with this problem, several investigations have been carried out. Contrast of indiscriminable colors in an input image can be improved by color conversion. In this report, we focus on the dichromacy, which is a kind of color deficiencies, and propose the color conversion method for it. In the proposed method, the contrast improvement is realized by modification of the lightness components in dichromacy. There are the protanopia, deuteranopia, and tritanopia in dichromacy. Then, we devised our method to cope with those color visions.

View full abstract

Lachman test is a manual test for examining the anterior cruciate ligament (ACL). An examiner examines the ACL by feeling finger stress. Some measurement device for Lachman test have been developed. However, they aim to analyze not the stress but a motion of the knee joint. Therefore, we set the goal of this study to develop a measurement system for finger stress with force sensor during Lachman test. In the experiment, one knee of a male without any injury was measured. In the results, the finger stress was observed by finger force sensor. As a result, we could detect the finger stress by the developed system.

View full abstract

In systems like chemical plants or circulatory systems, failures of piping, monitoring sensors or valves for control causes serious problems. These failures can be prevented by the increase in sensors and operators for condition monitoring. However, since the increase in cost is required by adding sensors and operators, it is not easy to realize. In this paper, a technique of diagnosing target systems is proposed by using a fuzzified neural network which is trained with time-series data with reliability grades which are given beforehand by domain experts. Our proposed technique makes us determine easily the state of the target systems because the state of a target system is determined based on the fuzzy output from the trained fuzzified neural network. From results of computer simulations, our proposed technique is flexibly applicable to various types of systems by considering some parameters for failure determination of target systems.

View full abstract

Recently, e-learning systems using computers and the Internet have been widely used with the spread of the personal computer to schools, universities and etc. An e-learning system has the advantage that a user can study with his pace and understanding level. However, it has the disadvantage that it is difficult for a user to keep the learning will well, and that the teacher can not visually catch the learner's situation. The purpose of this study is to compose the system that keeps learner's concentration level by taking pictures of the learner from two directions (front and side) with Web cameras. We judge whether the learner is studying well or not by classifying the pictures with the Adaboost and neural networks. Moreover, in order to show the effectiveness of these functions, we perform experiments to examine the accuracy of each judgment and we discuss the results.

View full abstract

A facility location problem is to find a location minimizing the sum of distances between demand points and facility points. If the optimal point should be selected from among the demand points, the problem is called the p-median problem. If the optimal point can be selected arbitrarily from a given domain, the problem is called the Weber problem. The p-median problem is a kind of combinatorial optimization problem. In this paper, a new heuristic algorithm for the p-median problem is proposed. Usually, enumeration methods should be avoided in case of such combinatorial optimization because the combinatorial explosion arises. It is, however, applicable for this problem if the nearest demand points to the optimal point can be specified. Proposed algorithm uses fuzzy c-means clustering for selecting some demand points considered to be near to the optimal points. Further, partial enumeration is done from selected points. In addition, the entropy maximization principle is used for adjusting locations of points for effective enumeration. There are some existing algorithms for Weber problem. This paper gives comparisons with these existing algorithms. In particular, the proposed algorithm has superiority over these existing algorithms for middle size p-median problems.

View full abstract

In recent years, various product regulations are developed aimed for environmental preservation or other social rationality. The companies selling any equipment or products are required to be pursuant to the established rules. For many product groupings, the regulations are relaxed using a framework named {\it Phase-In Method}. A product specification is changed one or more times a year, and each company pays for each change. There are two kinds of change of product specification, major change and minor change which costs are smaller than the major change. The goal of this paper is finding a solution minimizing the total cost consisting of the product specifications and pursuant to the rules using the tabu search.

View full abstract

Urban district heating and cooling (DHC) systems operate large freezers, heat exchangers, and boilers to stably and economically supply hot and cold water, steam etc., based on customers demand. We formulate an operation-planning problem as a nonlinear integer programming problem for an actual DHC plant. To reflect actual decision making appropriately, we incorporate contract-violation penalties into the running cost consisting of fuel and arrangements expenses. Since a yearly operation plan is necessary for check whether the minimum gas consumption contract is fulfilled or not, we need to solve long-term operation-planning problems. To fast and approximately solve long-term operation-planning problems, we propose a decomposition approach using coarse (monthly) approximate operation-planning problems.

View full abstract

In this paper, we focus on emotion recognition using prosodic features in speech. It can be occurred that prosodic features on the utterances with two different emotions may show similarity. Therefore, distributions of each emotion are overlapping, so that the accuracy of emotion recognition goes down. This paper proposes the emotion recognition technique which can distinguish each emotion category in overlapped distributions. In dialogues, emotion changes cause the change of utterance expression. So it assumes that prosodic features change remarkably on the emotion transition. In proposed technique, to recognize emotion in overlapped distributions of emotion, we put remarkable changes of prosodic features on emotion transition to use. In the experiment, we treated 4 emotion transitions; "Interest to Sympathy", "Interest to Joy", "Interest to Doubt" and "Interest to Antipathy". We calculated the differences on prosodic feature between before and after the emotion transitions, and drew histograms of differences of prosodic features. We show that there are differences among emotion transitions in points of kinds of changed features and their amount of change. Moreover, we show that there are significant differences statistically among the amount of change of prosodic features.

View full abstract

In this paper, we propose an emotion generation system based on MaC model using neural networks. In the proposed system, the chaotic neural network and the Boltzmann machine are used in the Emotion Generator of the MaC model. In the Boltzmann machine, the plural pattern can be recalled stochastically for the same input. The proposed system makes use of this property in order to generate different emotions for same external input. On the other hand, the chaotic neural network can recall the training patterns dynamically based on external input and the history. The proposed system makes use of this property in order to generate emotions based on its history. We carried out computer experiments, and confirmed that the proposed system can realize autonomous and human-like emotion generation. Moreover, we confirmed that the robot based on the proposed system can decide actions as similar as in the robot operated manually.

View full abstract

The term biological motion is often used by researchers studying the patterns of movement generated by living forms and sport forms. Biological motion data are acquired using 3D motion capture system. However, 3D motion cature system is too expensive. In this paper, we proposed human-emotion discrimination system using biological motion or one acceleration sencor. We study the method of a feature extraction and a pattern classifier to discriminate emotions of human walking style. The extracted feature and pattern classifier are evaluated from the discriminated result of the given walking data which identified four types of categories, Anger, Sad, Joy, and Neutral. We tested our proposed system using the computer simulation.

View full abstract

It is widely believed that there will be communication robots in the future working in our daily life, and researchers have started to study human-robot interaction. However, there are several factors influencing people's acceptance of these robots. This paper presents our findings on human psychological factors influencing communication behaviors toward robots, in particular, attitudes and anxiety toward robots, and then discusses about their implications on social acceptance of robots.

View full abstract

It is known that a good combination of mataheuristics with other optimization techniques can significantly improve the precision of solutions. In this paper, we propose a metaheuristic approach, namely Memetic Algorithm (MA), based on Tabu Search (TS) for the NP-hard k-minimum spanning tree problems. For an edge-weighted graph G=(V, E), to find a subtree in G with exactly k (k < |V|-1) edges, such that the sum of the weights is minimal. The genetic operator of our algorithm is based on an existing dynamic programming algorithm for finding an optimal subtree in a given tree, and the local search is based on an improved TS from the literature. The results of some numerical experiments show that MA has a good synergy with TS, when MA uses TS as individual steps for diversification and local optimization by agents.

View full abstract

In multimodal function optimization problems, which have multiple global optimal solutions, it is important to find all the global optima in one trial. Several species-based methods have been proposed to search for multiple global optima. In the methods, speciation mechanism is used for dividing individuals into some species, and each species searches a solution locally. The species-radius is used for the determination of the species' territory. But the parameter setting of the radius is dependent on objective functions. Therefore, it is difficult to determine the appropriate radius. To solve the problem, we propose a new speciation method using proximity graphs instead of the species radius. The proximity graph for individuals is constructed based on the concept of neighborhood in the graph theory. In our proposed method, Differential Evolution is used as the optimization algorithm. Moreover, several kinds of proximity graphs, relative neighborhood graph, Gabriel graph, beta-skeleton graph and so on, can be adopt for the speciation. We applied our proposed methods to several test functions and examined their effectiveness.

View full abstract

This paper studies about which games in extensive form are played by multiple players who make decisions in pre-decided order. The sub-game perfect equilibrium is known as a concepts of which explain or predict experimental data of human behavior of many games in extensive form. However, some results of laboratory experiments using human experiments indicate that the theoretical prediction do not always match the human behavior. A centipede game is one of such games. In many theoretical models, it is assumed that the players are rational and they can discriminate slight differences of payoffs between distinct strategies. In the real world, it is well known that the decision making mechanism of human are not only rational but also heuristic. In this paper, human subjects behavior in the centipede games are analyzed by a simulation system using artificial agents.

View full abstract

McCarthy (1960) defined the marketing strategy to be characterized by 4P; the Price, the Product, the Promotion and the Place. In general, to find the good marketing strategy in the real market requires substantial time and monetary cost, because of the analysis are conducted through the queationary investigations, approval sales, or other several experiments. In this paper, an artificial market system basing on adaptive artificial multi-agent system is constructed for effective marketing strategy analysis focusing on the effect of the trend to the marketing strategy.

View full abstract

Patients with delirium have a higher risk of falling and losing their lives. Therefore, delirium is known to be one of the factors that has a bad influence not only on treatment and nursing but also on the patient's vital prognosis. However, no specific therapeutic agents for delirium or prophylaxes against it have been developed yet. Besides, it has been reported that delirium is one of the mental states that is tend to be passed over. Accordingly, it is very important to develop the staffs' observation skills and to establish the nursing care system in order not to overlook delirium. In this study, we interviewed the clinical nurses working at Kansai Medical University Takii Hospital in a group setting. They are in contact on a routine basis with patients with delirium. Their spontaneous utterances about delirium are analyzed with the DEMATEL method. Their utterances are divided into two groups, that is, "causes of delirium" and "delirious patients' behavior". From each group, keywords and phrases are picked out and analyzed. Consequently, this study will reveal how the clinical nurses feel about delirium and delirious patients.

View full abstract

When children look for a book and wonder what to read, they usually ask adults some recommendation about a suitable one. In many cases, librarians or teachers play an important role in such cases. In other cases, parents or adults close to the children may give a recommendation for some books. On the other hand, the objective of the OPAC in libraries is to enhance children's search of books by themselves. Nevertheless, it is not easy for children to use this system. In this study, we aim at constructing the bibliographic search system that is easy for children to use. For this purpose, we studied situations where children used the book search engine: "Adventure in the Sea of Books," which we had developed for "International Institute for Children's Literature, Osaka" in 2002. This search engine is equipped with "Subject Heading Lists for Children" that we uniquely developed for this system. It is also installed with the book search function, through which users can search a book by its title, author, or any word they come up with. This study makes a comparison between the fixed keywords in "Subject Heading Lists for Children" and the words which users freely entered by themselves. This comparison shows what kind of keywords children really need, which will lead to the development of more user-friendly book search systems.

View full abstract

Rough set theory was proposed by Z.Pawlak in 1982. This theory can mine knowledge through a decision rule from a database, a web base, a set and so on. The decision rule is used for data analysis as well. And we can reason for an unknown object using the decision rule. In this paper, we apply the rough set theory to analysis of time series data. We acquire knowledge for tick-wise price fluctuations.

View full abstract

The objective of economic analysis is to interpret the past, present or future economic state by analyzing economical data. In many cases an economic analysis is pursued based on the time-series data or is analyzed the structure of the target system by multivariate data. Time-series analysis and regression model are central tools to analyze economics data. Nevertheless, economic systems are a complex system resulted from human behaviors and related to many factors. When the system includes much uncertainty such as ones of human behaviors, it is better to employ methodologies of a fuzzy system in the analysis. In this paper, we analyze Japanese economics by possibilistic forecasting model.

View full abstract

This paper deals with the Vender Managed Inventory (VMI) supply chain. The VMI is a production and inventory control strategy where stock positions are well known throughout the supply chain for the purposes of setting production and distribution targets. In this paper, we will focus our attention on the control strategy based on a VMI model. In addition, simulation results will be provided in order to verify the effectiveness of the control strategy in the presentation.

View full abstract

In this paper, we try to model for 'Naiji System' which is a unique corporation between a maker and suppliers in Japan. We propose Mass Customization Production Planning & Management System (MCPS) based on unfulfilled-order-rate by using Advance Demand Information, which is called 'Naiji'. This model is formulated as a nonlinear stochastic programming problem which minimizes the sum of production cost and inventory holding cost subject to the set of probabilistic constraint and some linear production constraints.We propose that (i)the strictly defined indicator of the unfulfilled-order-rate (SOn), (ii) the upper bound to estimate the unfulfilled-order-rate more strictly(SOn(min)), (iii)the indicator of the unfulfilled-order-rate in the case that inventory is independent between each periods and we analyze performance of 3 indicator. We confirm that inventory becomes 7-12% decreases by using the indicator of SOn,comparing with SOn(0). Finally an availability of the indicator and of the procedure is shown.

View full abstract

Traditionally, a x_bar control chart has been used together with s or R control chart to monitor a state of the production process. But there is such a fault that it is hard to interpret the connection of behavior between mean and dispersion in the process because respective characteristics are monitored by an individual method in parallel. A (x_bar,s) simultaneous control chart has been proposed in order to enable the process manager to monitor the change in the process mean, process dispersion, or both. Also, identifying which of the process parameters is responsible for the out-of-control condition of the process is one of the important issues in the process management. Especially, it is an important task in the (x_bar,s) simultaneous control chart because the observations of some parameters are connected. In the traditional literature, the multiple decision method based on the statistical hypothesis tests and the decomposition method of the statistic by the perspective of Kullback-Leibler information are proposed for the purpose of identifying the parameter responsible for the out-of-control condition. In this paper, we propose how to identify the parameter responsible for the out-of-control condition using Akaike information criterion. Then, the validity of the proposed technique is shown through the numerical experiment.

View full abstract

We developed the management system for Sakematsuri QR code rally. It aimed at the cost reduction and the user process tracking. This system was used in the Sakematsuri of Higashihiroshima City last year.As a result, User's route was able to be traced by the access log analysis.

View full abstract

Recently, food service industry including restaurants and fast-food chains has been explosive growth, so the reduction of labor cost and the devices to promote purchasing rate and sales figures are extremely important task for each store. To resolve these issues, some inventory management systems with Point-of-Sales (POS) have been previously developed. These systems make the improvement of service quality and sales possible by reducing the labor cost of staffs and the opportunity loss on peak time. However they need expensive specialized equipments and maintenance cost, so only large chains can introduce such systems. The purpose of this paper is to improve operational efficiency of restaurants regardless of their scale. Especially this paper develops inventory management system cooperating with POS and automatic menu-ordering capabilities with Web technology. This paper aims the provision of services cheaply and efficiently by using personal potable digital devices with Wi-Fi and touch-sensitive panel as menu-ordering device. The system can visually-present sales information depending on seasonal movements, goods, and customers, so it might be possible to support the reduction of disposal loss and inventory amounts.

View full abstract

Recently, how to take communications has changed by the development of the information society. Communications by mail and the blog, etc. increase. Therefore, interpersonal relationship management skill deficiency is problem. In this study, the communications means between students is investigated by using the Web questionnaire. The correlation between students is analyzed from the questionnaire data, and the connection degree is shown in figure. The ideal method of communications between the teacher and the student is thought based on those analysis result and figure. Moreover, it is thought that it leads to the promotion of the educational activity by comparing the analysis result with the result.

View full abstract

We apply the Network DEA technique to the management efficiency improvement in a public museums from medium and long term viewpoints. When an evaluation is performed based on it, we analyze the environment factors to become non-efficiency with Tobit model. Furthermore, we reevaluate the management efficiency of the public museums by correcting the result.

View full abstract

In recent years, the decision-making process in case a consumer chooses goods is also complicated because diversification of taste and goods became a various kind. Especially the decision-making process in buyer behavior has been changing a lot with change of a time. When the manager sets up the managerial policy, it is important to analyze the process of the decision making of the selection based on consumer's purchase behavior. The selection in the consumer changes by the interaction with environment, and, as a result, the criteria of each user is diversified. When the consumer selects the action based on the criteria, it is thought that the randomness is influenced by fuzzyness in the transition history of sense of values and the uncertainty of the object recognition. In this paper, we study discusses the decision-making model concerning the selection of the commodity in consumer's purchase behavior and the mechanism.

View full abstract

The clonal selection principle is used to explain the basic features of an adaptive immune response to a antigenic stimulus. It established the idea that only those cells that recognize the antigens are selected to proliferate and differentiate. This paper explains a computational implementation of the clonal selection principle that explicitly takes into account the affinity maturation of the immune response. The clonal selection algorithm to find the antigen by the combination of Immunological Memory cells and the proliferation of immune cells. This paper tries to classify the medical database of Coronary Heart Disease databases and reports the computational results for 4 kinds of training datasets.

View full abstract

Bladder cancer is the most common urinary cancer. Even after surgical operation of the bladder cancer, 50% recur within two years. With the additional use of carcinostatic substance the recurrence rate is lowered to 20-30%. However, post-operation recurrence prediction has not yet been established for bladder cancer, increasing the risk for patients. The prediction problem is further complicated by the fuzziness of medical data. Previous research using artificial neural network reported 71.6% correct classification of cancer relapse. This paper reports on applying ensemble learning of artificial neural networks to improve the classification results. Bagging using multiple data sets were used for the ensemble method.

View full abstract

In this paper we applied cooperative learning artificial neural networks (ANN) using particle swarm optimization (PSO) for paddy field classification from remote sensing data. Remote sensing is commonly used for monitoring of change in the earth's environment. In the field of land-cover classification, classification methods using moderate resolution satellite sensor data is actively researched. Paddy field classification strongly depends on temporal changes that can differ among regions and countries. Therefore it has been difficult to generate a common paddy field classifier using previous statistical methods applicable to different countries and regions. Our aim is to automatically generate a paddy field classifier in order to create localized classifiers for the required country and region. In our previous research using ANN and spatiotemporal satellite sensor data, the proposed classifier yielded 90.8% correct classification rate. On the basis of these results, we applied a cooperative learning method using PSO as the global search method and ANN as the local search method for improving the classifier accuracy.

View full abstract

Our aim is to analyze the rhetorical structure of an online manual with Microsoft Office Word (called Word) automatically for extracting the important texts in it. The rules of rhetorical structure, which are generated from the feature of the texts in 25 titles with Word 2000 on the operation on character strings, are used to automatically obtain the rhetorical structure of the texts. We evaluated the precision of the rules by comparing the automatic analysis to the manual one. The three kinds of texts are used for evaluation: 1) the texts in 25 titles used for generating the rules, 2) the texts other than the texts for generating the rules, 3) the texts in 25 titles with Word 2007.

View full abstract

The process of creating a figure takes a long time. In this paper, we present our studies to realize an interactive drawing system, which interacts with a user through natural language, and gathers information to create a figure. First, we conduct a subject experiment to collect texts that describe drawing procedures, and construct a prototype drawing system based on these texts. Second, in order to improve practical utility of the prototype system, we conduct another experiment to collect dialogues between subjects who play roles of a user and a drawing system. By analyzing these dialogues, we sort out key functions for a practical interactive drawing system.

View full abstract

Recently as digital cameras and web cameras have been commonly used in our everyday lives, we can easily obtain quite a few movies. However, it is difficult for us to find a particular part of a movie. Considering this, in this paper we propose a method to verbalize human behaviors in a movie by observing how a human being interacts with objects in a particular space with image recognition technology.

View full abstract