Many factors are required to demonstrate the urban activity for evaluation of city planning policy in the local city. The multi-agent model is proposed to be established concerning with many urban activities and interaction between agents. In addition to the fundamental purpose of regional development in city development policy evaluation, the evaluation method of city development is discussed connecting with the problems in local city. The multi-agent model is proposed as an artificial society involving measurement mechanism of environmental load. The citizen agents and commercial store agents are assumed in the complex system model. The fundamental formulation of the agent activities is discussed. Furthermore, the evolutional process for urban activities is additionally mentioned. As the artificial society as summary of urban citizens is established, comprehensive evaluation of city planning policy becomes available and the sustainable direction of city planning can be shown with observation of future change of society.

View full abstract

The number of deteriorated infrastructure has increased dramatically in Japan. As a means of inspection for existing concrete structures, visual inspections are generally carried out to draw a sketch of cracks and measure crack width. Crack width is an important data in the diagnosis for the safety of concrete structures. The crack width measurement is done manually by inspectors using visual crack width gauge. The inspection takes an incredibly long time. Therefore, there is a significant demand for advanced tools for visual inspection. Recently, digital image processing technologies have been successfully applied in different industries. This paper describes an approach for automatically measuring crack width on concrete structures using image processing.

View full abstract

A purpose of this study is to develop the idea generation support system giving "discovery" to the designer for the facade design of an office building. This study analyzes the form elements constituting a facade and develops the idea generation support system reducing of the user's psychological burden using IDE. 58 form elements are extracted by collecting the image of an existing office building and classifying. 58 form elements are replaced to the individual's vector of IDE, and are incorporated into IDE algorithm. 3D image of facade by IDE algorithm is expressed by VRML. The verification of this system was performed by 10 students learning architecture. In the system evaluations, there are small differences by evaluator's personal backgrounds, but the classification of form elements is proper because the exciting designs are generated for about 90% of all the trials and about half trials satisfied the end condition. The number of data is a few, but this system is effective because this system can offer some exciting designs to all evaluators at a comparatively early generation.

View full abstract

Recently, technology for producing 3D model space from optical satellite is paid to attention as a technique that takes the place of photogrammetry.But it is known to lack stability because of current photogrammetry.Then,this research tried the production of 3D model space with satellites images.

View full abstract

In order to establish a rational bridge maintenance program, it is inevitable to determine the deterioration curves of various components of bridges. However, it is difficult to identify appropriate deterioration curves because of various uncertainties involved in the prediction process and environment evaluation. Furthermore, it is necessary to take into account the effects of natural hazards. Under the situation, although the existing bridges should be kept in a safety condition through inspection and maintenance works, it is not easy to maintain them in a satisfactory level, because of the financial constraints. In this study, an attempt is made to develop an optimal maintenance planning system. In the proposed system, an optimal maintenance scheduling is obtained with Genetic Algorithm and Monte Carlo Simulation, in which the deterioration curve can be updated by using the data given from the structural health monitoring or periodical inspection. Several numerical examples are presented to demonstrate the applicability and efficiency of the proposed method.

View full abstract

In architectural planning stage, multi-agent system based on human behavior models is considered to be effective in order to estimate space characteristics of objective buildings such as efficiency, amenity, safety, and so on. In multi-agent simulations, at first, human behavior rules are set up, and the efficiency of the objective space is estimated by observed results on agents' behaviors with employed rules. However, it is difficult to configure human behavior rules, because actual human behaviors are complex. Therefore, acquisition methods of human behavior rules in accordance with objective spaces and/or issues are considered to be an important problem. In this paper, aiming to develop adaptable agent models, artificial neural network (ANN) is employed as acquisition and control methods of agents' behaviors. As for training methods of ANN, genetic algorithm (GA) is employed in order to determine optimal weights and biases of ANN. As for objective spaces, passage spaces with complicated shapes are employed. Based on simulation results, the effectiveness of acquired behavior rules by GA, the generalization capabilities of ANN, and the effectiveness of multi-agent simulations are discussed. Furthermore, possibilities on acquisition methods of adaptable human behavior models are also discussed.

View full abstract

The global warming problem by greenhouse gas emission has become more serious. Also in the transportation section, construction of the sustainable transportation system in environment has been an important subject. Not only improvement of service level of modes except private vehicle, but also consciousness for environmental problem (eco consciousness) of individual trip maker is important for eco-commuting promotion. Eco consciousness would be changed by influence of other people. The multi agent simulation system for planning of eco-commuting promotion is developed with social interaction model for eco consciousness in the study. Modal shift is modeled by considering the result of a transport social experiment for eco-commuting promotion as reference. Influence of promotion plan such as improvement of service level of public transport, promotion of individual eco consciousness and strengthening interaction of individuals can be analyzed with the proposed model. Finally, it can be concluded that the proposed multi agent simulation with social interaction for eco consciousness is useful for planning of eco-commuting promotion.

View full abstract

Global warming and its mitigation measures such as GHG (Greenhouse Gas) reduction have been extensively discussed all over the world. The transport sector occupied about 20% within the GHG of Japan. Countermeasures in the road traffic are urgent, since about 90% of the GHG emission of the transport sector is emitted from automobiles. For the effective reduction of the GHG such as CO2, quantitative analysis is necessary. Though the emission from the individual automobile depends on various factors such as engine rotational frequency, speed, and acceleration, that relation is complicated. Therefore, there is a limit on the modeling by the regression analysis. Then, the model which estimates the CO2 emission from the individual automobile using fuzzy neural network is developed in this study. By including this model in the traffic simulation, the estimation of CO2 emission becomes possible. Therefore, it is possible to examine traffic countermeasures considering the environmental effect.

View full abstract

This paper describes on desk layout in an office considering moving distance and human relationship when working groups exist under the condition that some need to collaborate each other and others do not. Each desk in the office, the agent recognizes desks that belong to working groups to collaborate with and ones occupied by people who don't have good human relationships. First it chooses randomly one of cells in Moore neighborhood and calculates summation of distances to all desks to collaborate with. If the value is smaller than one in the former step, it moves into the cell. However, if distance to the desks with unfavorable relationships is shortened, it stops to move and stays in the former position. When this operation applied to all desks, the first step is completed. Finally several desk layouts with certain rationality are obtained although they seem to be random and look different from ones arranged normally in the office.

View full abstract

The most appraisal theories on affection or emotion were a widespread agreement in terms of evaluating beneficial / harmful mean about substance perceived scenes. However, structures and process were not necessarily match.

View full abstract

We proposed some methods to understand partner's emotion from contents of utterances and to express facial expression of computer agent according to agent's emotion. However, if we construct such methods ad hoc, its model will apart from process of real mind. In this paper, we reconstruct our mental model based on discussion in philosophy of mind. In philosophical field, there are two approaches, monism which deal mind and matter with the same one and dualism which says mental phenomena are non-physical. In this research, we use "interactionist dualism." Our mental model are separated into physical network which deal with deductive reasoning, emotion network which calculates aroused emotions, and mood network which transits the mental state of the computer agent. Furthermore, some nodes of mental state and physical state are connected according to causal relationship and it outputs facial expression and changes action of the agent.

View full abstract

Research on Emotions within Artificial Intelligence and related fields has flourished rapidly through several years. Unfortunately, in much research the contextuality of emotions is disregarded. In this paper we argue, that recognizing emotions without recognizing their context is incomplete and cannot be sufficient for real-world applications. We present logical underpinnings of this claim and describe some consequences of disregarding the context of emotions. We also present our approach, in which this context is considered and describe some of the first experiments performed in this matter. The paper is finalized with a discussion on future development and applications of context processing within Affective Computing.

View full abstract

This paper proposes an illumination system reflecting character's feelings and moods of scenes in a picture story show. Inputs to the system are degrees of character's feelings represented by the value in [0, 1] and moods of scenes represented by 12 pairs of adjectives. The relationship between character's feelings and parameters values of lighting, and that between moods of scenes and parameters values of lighting are represented by neural network models. This system illuminates pictures using LED according to obtained parameters values of lighting. Two types of illuminations are considered in this paper: One is illumination on picture reflecting moods of story scenes, and the other is spotlightic illumination reflecting character's feelings. This paper also describes subjects experiments for the confirmation of the validity of the proposed system.

View full abstract

IDS(Ink Drop Spread) method is a modeling technique that copy the method of handling information in the human brain based on the concept of the soft computing.This technique replaces the characteristic of the object system with two or more two dimension pattern information.That is simple and confusion, transforms those patterns by study, combines them finally by a fuzzy technique, and does the modeling of the object system.The object system was replaced with pattern two capitation crack dimensions information in the IDS method at each input. However, a segmented region suitable in each object system is calculated in other modeling techniques and modeling is done.Then, we examine the performance when two or more approach to decompositions are prepared, and each technique is applied to nonlinear function approximation.

View full abstract

One of the most important designs for a lot of machine learning methods is the determination of the similarity between instances. Especially the kernel matrix, which is also known as the Gram matrix, plays a central role in the kernel machines such as support vector machine. The simplest design of similarity function is to use the distances between instances or the Gaussian function based on them. It is easy to learn the model when the data distribution follows their label, in which the instances with same label are allocated near and those with different label are allocated far. However, when the data distribution is non-linear, it becomes difficult. This paper discusses the inner products of 2 non-orthogonal basis vectors and proposes the similarity between instances. This paper also proposes a space folding model for machine learning based on the proposed similarity. This paper applies the proposed method to pattern recognition problem and shows the effectiveness of it.

View full abstract

The bilateral filter is an edge-preserving smoothing filter to reduce the superimposed noise on a signal. The tuning of smoothing parameters in the bilateral filter using an input signal is needed to produce a good reconstructed one. The difference between the input and reconstructed signals is regarded as a noise component. If the similarity between the estimated noise distribution and the assumed noise distribution model is high, it will be expected that a good reconstructed signal is given. In the proposed method, the similarity is calculated using the Hellinger distance and the smoothing parameters in the bilateral filter are tuned based on it.

View full abstract

This paper presents a relaxation approach to analysis and design of discrete Takagi-Sugeno fuzzy systems. Our relaxation approach does not require that the difference of a given potential function is always negative, i.e., that a potential function V monotonically decreases with respect to time t. We derive relaxed stability and stabilizability conditions that a potential function V is not always required to monotonically decrease. The derived conditions are represented in terms of linear matrix inequalities. To illustrate the validity and applicability of the proposed approach, two analysis and design examples are provided. The examples demonstrate the viability of our proposed approach.

View full abstract

There is a large number of researches that focus on the fuzzy control system by using the so-called T-S fuzzy model, which represents the system to be controlled in terms of state-space representation in some divided sub-fuzzy spaces. However, few studies have paid attention on the reconstruction error between the system to be controlled and its T-S fuzzy model. In this paper, an adaptive fuzzy controller is proposed in which fuzzy approximator is used to deal with the reconstruction error.

View full abstract

This paper discusses sampled-data control design for fuzzy systems by the first-order hold input. When we use digital devices to control systems, we need a sampled-data control input. Here, we employ the first-order hold sampled-data control input. Since the first-order hold input can be regarded as a delayed signal, the closed-loop system with the first-order hold state feedback control becomes a system with time-varying delay in state. We first give a sufficient condition for the stability of the closed-loop system in terms of linear matrix inequalities(LMIs). Based on such a stability condition, we propose a design method of the first-order hold sampled-data state feedback controller.

View full abstract

The partial role of J-STAGE is a conference management system which has been designed to help programme chairs to cope with the complexity of the refereeing process. FSS uses the system for the sophisticated and flexible management of the access of PC members. However, recent FSSs provide many kinds of organized sessions. Many session organizers cannot access J-STAGE system, because of the security and the management. Therefore, we developed the WebDB system to realize a smooth communication between PC members and session organizers.

View full abstract

Elderly and Young person groups were asked to estimate the degree of relation between the independence level of 1-80 year old and 5 developmental estimation item groups in this study. In both groups, mean scores of 5 developmental estimation item groups of each age were changed by various patterns. Some different types were shown in estimation patterns of both groups. The appearance ratio of the increase-decrease type of elderly group was higher than that of the young person group. The appearance ratio of the increase-constancy type and the decrease-increase type of young person group was higher than elderly group. The difference of the appearance ratio of each type between 2 groups was significant. Difference of recognition of the developmental change and independence level between elderly and young person group and influence concerning difference of estimation method are discussed.

View full abstract

This paper analyzes factors to affect escape from web sites, taking site properties and personal features into account. Down load time and layout conditions are adopted as site properties, and interest, stress, and Type A behavior pattern as personal features. Frequency of escape is investigated by experiments, and personal features and evaluation for web sites are collected by questionnaires. The relationship between scores of the Type A behavior pattern test and frequency of escape is analyzed by the quantification method II. It is shown that subjects having a high score of the Type A behavior pattern test tend to escape from a web site if they are less interested in it, if down load time is longer, and if the layout condition is better. Also, there seems to be the possibility that relief of stress causes escape from web sites.

View full abstract

Companies often employ questionnaire(s) in order to design marketing strategies or to grasp the trends. Questionnaire data is a large and multi-dimensional data. The authors have analyzed based on visualized results using dimensional reduction and visualization methods such as PCA (Principal Component Analysis) and MDS (Multi-Dimensional Scaling method.) However, if it does not consider the questions which many respondents thought similar meanings, there is a possibility that the characteristics of respondents can not be found out because visualization results are more emphasized by the similar questions than the other ones. This paper defines the distance between questions which represents the similarity of questions, and proposes a method to derive questions having similar meanings for respondents and contract them based on the distance between questions. This paper applies the proposed method to an actual questionnaire data and shows that the features of some groups of respondents can be found while it is difficult in the conventional visualized result.

View full abstract

Recently, companies often carry out questionnaires. These questionnaires often have questions which require respondents to answer by free description. It is, however, inefficient for the analysis of these texts to read whole data for getting outline or classifying them. The authors have proposed and introduced the HK Graph (Hierarchical Keyword Graph) which is a support tool for text mining. HK Graph can visualize the relationship among attributes and words with hierarchical network structure based on the co-occurrence information. However, the result of HK Graph gives us the relationship between words in the target text data, which is not enough to grasp the outline of the texts and extract features of them from only the visualized result. This paper considers the similarities of words and groups words based on their "concept," and tries to feed back them into the visualization result.

View full abstract

A multiple (dis)similarity aggregation framework is proposed for robust selection of suitable measures for a query. Ranges of all measures are normalized and transformed into a single similarity measure by re-scoring them. Precision levels and computational costs on the Corel database are compared to a system that uses a single measure. The combination of Minkowski distance and cosine similarity achieves the best result in the sense that its computational overhead is only 0.1 seconds when 100,000 images and 1,000 visual words are used. Due to the robustness of measures in this combination, applications include retrieval employing large number of image categories in a database.

View full abstract

This paper reports on algorithm education using logical data structure and the core image of an algorithm. In algorithm education, the ideas of logical data structure and the core image of an algorithm aren't usually used, but it can be effective for the deep understanding of algorithm to use them. Some practical samples using them are given.

View full abstract

The measurement control of the machine is the theme which is easy to understand for the learner of programming to discuss numerical value data. We propose the programming guidance which used a measurement control subject for Numerical value processing and Graphics.

View full abstract

In the teaching of informatics, it is especially important that a careful selection of appropriate teaching materials and method be made. The author has been endeavoring to include both traditional teaching and ICT in teaching materials and method in recent years. This paper discusses the selection of teaching materials and method from both traditional teaching and ICT. A proposal is made for the teaching of hybrid informatics by including traditional teaching and ICT.

View full abstract

We propose that the consciousness is produced from the neural networks driven by the outer and inner stimuli, besides we present the evotional study of feelings. The feelings may be believed as strong and changeless part of the consciousness, but we represent that the feeling themselves will evolve into several another feelings.

View full abstract

The technology of information networks is necessary for our life. Information network cabling professionals perform important skills in order to build information network infrastructures.We recognize the importance of improvement in vocational training. An information network cabling professional is required to have skills to maintain infrastructure development and execute cabling such as optical fiber. I introduce an effective example of vocational training in the field of information network cabling by using the system of vocational qualification.

View full abstract

It is known that model-based fuzzy control systems can result in the solvable problem of LMI which satisfies existence condition of Lyapunov function, namely stability condition of control systems. Such LMI condition may become conservative, and it is important to propose the design method of fuzzy control systems that more relaxed stability conditions are obtained. In this study, the output feedback type regulator is constituted for controlled object expressed by T-S fuzzy model involving disturbances, and boundedness condition of the time-varying linear inhomogeneous differential equation systems of the closed-loop systems is proposed. The validity of control system design method proposed from the numerical example is verified.

View full abstract

This paper studies coordination games in endogenous networks. A strategy of a player is a combination of choice of actions in the coordination game and choices consisting of partners for playing the game. The choice of partner by a player means that the player chooses the partner of the link formation. The game is played by a pair of two players if there exists a link between the two players or there exists a path from one player to the other. We examine the formation of interaction in a network. In our mathematical model, it is assumed that any two players who have a path whose length is 2 or have a link interaction to each other, i.e.,they play the coordination game. In this paper, we indicate the strict Nash equilibrium of the coordination games played in endogenous network.

View full abstract

We deal with equilibrium selection in large populations of agents where two agents are selected for playing a coordination game. We consider two simulation models corresponding to the Young model (1993) for a single population and the Bhaskar and Vega-Redondo model (2004) for multiple populations with migration. In each of the two models, we develop two decision mechanisms; one is based on best response and the other employs adaptive behavior, and we compare two sets of actions of the artificial agents in the best response and the adaptive behavior models by using the simulation systems.

View full abstract

Genetic programming (GP) is one of the techniques for evolutionary optimization of programs. Linear GP (LGP) has been proposed as an extended GP. Individuals, whose structure consists of a linear array, are represented by a sequence of low-level language such as an instruction code. In LGP, a fitness value is calculated by processing codes serially from head position and changing contents of the register. The instructions that individual has are referred to as an effective code only if the result of its calculation makes an impact on the fitness value, otherwise a non-effective code. It is possible to change the representation of individuals effectively by applying genetic operations to effective codes. In this paper, to seek a better solution within lower generations by intensification of search, we proposed a new genetic operation that inserts the effective piece of codes existing in better individuals to the other individuals. We performed an experiment for comparing standard LGP and LGP with the proposed genetic operation using benchmark problem. The result indicated the proposal method is useful.

View full abstract

We propose Genetic Programming(GP) with control nodes using conditional probability. In our method(GPCN_CP), each individual consists of some trees. In order to improve the efficiency, individuals in the next generation are generated by using genetic operations or conditional probability tables, where some individuals with high fitness values are used to evaluate conditional probability tables. We have implemented three methods, traditional GP, GP with control nodes(GPCN) and GPCN_CP, and these methods are applied to a garbage collector problem. Experimental results shows that GPCN_CP is more efficient than traditional GP and GPCN.

View full abstract

In this paper, we have implemented some incremental classifier based on ARTMAP and we have compared the performance of these methods. There are two modified methods of ARTMAP that can take into account the degree of importance of attributes, ARTMAP-AW and ARTMAP-MD. ARTMAP-AW defines categories as hyper ellipses and it can ignore attributes irrelevant to classification on the basis of the importance attributes. In ARTMAP-MD, Mahalanobis distance is used to define the distance between a category and a case, so that it take into account not only the importance of attributes but also correlations between attributes. Although experimental results show that ARTMAP-AW acquires better classification rules than ARTMAP-MD, we have pointed out the modification to improve the performance of ARTMAP-MD.

View full abstract

To find out the mechanism of brain activity during emotional elicitation is an important issue for understanding mental disorders including depression and anxiety disorders. Prefrontal cortex of the brain has received considerable attention because it is thought to have the important rule of emotion regulation. This presentation provides an overview of previous study on emotion and the brain, and the introduction of our research. In our research, near-infrared spectroscopy (NIRS) was used to measure prefrontal activities with healthy college students during emotion elicitation. NIRS is the noninvasive device that can mesure cerebral blood flow. The International Affective Picture System (IAPS; Lang et al., 1995) was used to elicit emotions. Fifteen negative, 15 neutral and 15 positive pictures of the IAPS were presented in a block design. Prefrontal activity was compared in each block.

View full abstract

Expectation is an adaptive function which promotes information processing by preparing behavioral, emotional, and physiological systems for upcoming environmental stimuli. However, how expectation of upcoming emotional stimuli modulates processing and evaluation of the stimuli itself is unclear. Thus, present study examined whether expectation of visual unpleasant pictures modulates processing of the pictures. To accomplish this purpose, we compared subjective and electrophysiological responses to unpleasant pictures preceded by erroneously-induced high estimation of unpleasantness about upcoming pictures with an estimation which corresponded to actual unpleasantness. As a result, subjective unpleasantness rating and an electrophysiological response were higher when participants highly estimated unpleasantness of upcoming pictures than they had practical estimation. Present results indicated that emotion can not only be elicited by bottom-up information processing of environmental stimuli but also modulated by top-down expectation of the stimuli.

View full abstract

This study examined the regulation of emotional expression in dyadic communication from the perspective of social skills. When strangers encounter one another, regulation of emotional expression is perceived, particularly in Japanese culture, as an adaptive communication style, part of social skills. Participants (n = 66) viewed a relaxing film for control. Then they viewed one of the films that evoked active and positive affect, deactive and positive affect, and neutral affect. They were paired off into dyads, and their 6-minute conversation was videotaped. As predicted, communication behavior was not influenced by affect in high social skills (n = 36). These results were discussed from the perspective of relationships between participants, and method of affect induction.

View full abstract

This paper builds hierarchical clustering methods using Formal Concept Analysis. It proposes algorithms of the single-link and complete-link methods through Formal Concept Analysis. An application of these new methods is illustrated by a simple example.

View full abstract

Recently the analysis method of Rough Sets came to be applied for the investigation for making design plans of the new products. So, it has been shown that the method of extracting Decision classes for Rough Sets were very important. Therefore, we suggested the new inferential method of Decision classes with a simple calculation method of Attribute Reduction in Variable Precision Rough Set Models proposed by Ziarko. Moreover, we suggested the new squeezing method of Decision classes in the conclusion of Decision table by the method of considering the simple calculation method of Attribute Reduction with Lower-approximation by Kudo and Murai.

View full abstract

Through the case with the electric shaver, we have considered and verified the effectiveness of the method of fixing the decision class by Variable precision rough sets proposed in this paper (No.1). In addition, we also considered how to apply it to the design concept making through the case with the digital camera. As a result, we have found that the proposal method was the effective and able to apply the design concept making.

View full abstract

It was reported that when subjects hesitated over choosing attractive one of a pair of stimuli, significant changes of their eye movements appeared. This paper, examining features of eye movements in such a situation by experiments, develops a model that infers degrees of hesitation from important movement parameters.

View full abstract

We verify the relation between physics parameter of robot and Human's Kanseiby factorial analysis for entertainment robot whose seems to be living any emotion.Because method of expressing behavior of robot in variegation is verified.We use to experiment a robot arm for body movement psychology.Because the robotic arm can do complex movement by an easy parameter setting,the operation of various patterns can be verified.Clarifying the relations of the emotion expression in a robot's behavior and user's Kansei,the robot can express feelings from complex behavior.

View full abstract

Arithmetically synthesized sentences are used for malicious purposes such as unsolicited commercial junk submission to web site. We study a problem to distinguish between natural and synthesized messages generated in Markov chains and show experimental results. Based on the difficulties of the problem, we consider a new application to CAPTCHA, a type of challenge-response test used in computing to ensure that the response is not generated by a computer.

View full abstract

The Shapley value has been proposed as an important index of fuzzy measures. The Shapley value can be interpreted as the following two representations: 1) an expected marginal contributions and 2) an cumulative value of dividends. In ordinary cases, these different two interpretations lead to the same notion, i.e., the Shapley value, while in some restricted domains, to different notions. In this paper, we show notions/indices which are led by two approaches based on above interpretations. Moreover, we suggest how and in what situations these indices should be applied.

View full abstract

-

View full abstract

Some types of Vector-valued Choquet integral models are proposed. For an input vector x = (x_1,..., x_n)^T, the output y = (y_1,..., y_m)^T is calculated by the m times (extended) Choquet integral with respect to the jth fuzzy measure mu_j. The vector-valued Choquet integral model is an extension of the product calculation between a matrix and a vector. The logical type vector-valued Choquet integral models enable to do classifications. If the sum of fuzzy measure values among fuzzy measures equal to 1 for all sets, the sum of output values is 1. Lastly, we propose a classification method with overlapping and unclassified degrees from ordinal classification by fuzzy rules.

View full abstract

We generalize Choquet Integrals with respect to two additive measure and using them approximate functionals defined on a finite set. We consider a product probability measure on the functional space, which actually is an Euclidian space. Then our approximations are some projection of the L2 space. Under this condition, we obtain that: the optimal approximation of a multi lenear Fuzzy integral is a multi lenar Fuzzy integral, and the optimal approximation of a Choquet integral is a different type integral.

View full abstract

The purpose of our research is to develop an algorithm which enables to obtain multi-class data representation by separating two independent components, namely, general model and individuality components. General model represents the common data structure for all classes, whereas the individuality component modifies the general model as the parameters for data generation of each class. The proposal method is consist of a higher-order nonlinear manifold learning algorithm and a linear subspace algorithm, thus, combinations of SOM2-PCA and SOM2-NMF. In this report the method is introduced as well as some preliminary results.

View full abstract