Journal of Advanced Computational Intelligence and Intelligent Informatics
Online ISSN : 1883-8014
Print ISSN : 1343-0130
ISSN-L : 1883-8014
18 巻 , 6 号
選択された号の論文の21件中1~21を表示しています
Regular Papers
  • Hiroshi Takahashi
    原稿種別: Paper
    2014 年 18 巻 6 号 p. 875-887
    発行日: 2014/11/20
    公開日: 2019/07/01
    ジャーナル 認証あり

    This paper examines the possibility of a new warning method that would increase drivers’ sensitivity to hazardous factors in the driving environment. The method is based on a visual warning mark in the peripheral vision, called an ambient warning. In this study, the use of ambient visual marks is investigated. These marks are soft visual warnings and lack officious interference with the task performed in the central vision field. Experimentation with a 27-inch monitor display resulted in decreased response times for detecting a flashing mark when an ambient mark was shown in advance. The results suggest that information observed in the peripheral vision field could help people drive more safely.

  • Woo-Jin Choi, John-Tark Lee
    原稿種別: Paper
    2014 年 18 巻 6 号 p. 888-895
    発行日: 2014/11/20
    公開日: 2019/07/01
    ジャーナル 認証あり

    Precise measurement of levels of liquids stored in tanks is essential for monitoring and predicting disasters by detecting leakages or arbitrary discharge of toxic materials. Therefore, tanks are typically equipped with a series of liquid level sensors. A magnetostrictive-type level sensor is composed of a waveguide, a current pulse interrogator, and a sensing coil for detecting reflective torsional signals caused by Wiedemann effect, which is the main principle of operation of magnetostrictive-type liquid level sensors. In order to implement a high-precision magnetostrictive-type liquid level measurement system, we used time–frequency analysis techniques such as wavelet transform (WT) to precisely detect the reflected signals. By using time–frequency analysis techniques such as short-time Fourier transform (STFT) and WT, a robust and precise liquid level measurement system can be implemented even in noisy environments.

  • Maslina Zolkepli, Fangyan Dong, Kaoru Hirota
    原稿種別: Paper
    2014 年 18 巻 6 号 p. 896-907
    発行日: 2014/11/20
    公開日: 2019/07/01
    ジャーナル 認証あり

    Bibliographic big data visualization method is proposed by incorporating a combination of fuzzy c-means clustering and the Newman-Girvan clustering algorithm, where clustered results are displayed in a network view by grouping objects with similar cluster memberships. As current bibliographic visualizations focus on the crisp relationship among data, fuzzy analysis and visualization may offer insights to bibliographic big data, enabling faster decision making by improving displayed information precision. The proposed method is applied to the DBLP citation network dataset. Results show that merging two clustering algorithms and visualization using fuzzy techniques enables the user to converge a few target papers within an average of 5 minutes from 1.5 million papers stored in the DBLP. Users targeted for the proposed method include researchers, educators, and students who hope to use real-world social and biological networks. The proposal is planned to be opened to the public through the Internet.

  • Hugang Han, Hak-Keung Lam
    原稿種別: Paper
    2014 年 18 巻 6 号 p. 908-917
    発行日: 2014/11/20
    公開日: 2019/07/01
    ジャーナル 認証あり

    This paper proposes a discrete sliding-mode controller for a class of nonlinear systems described by a T-S fuzzy model subject to modeling error, which may influence the system performance and the overall system stability. While most of existing literature treats the modeling error under the so-called parallel distributed compensation framework by using some norm-bounded matrices, the proposed control scheme in this paper integrates a feedback component, which mainly consists of fuzzy approximators to deal with the modeling error and an auxiliary component of the variable structure control with a sector to guarantee the global stability of the closed-loop system when the system state travels outside the sector. With the consideration of system stability, adaptive laws adjusting the parameters in the system are developed based on the Lyapunov synthesis approach. Finally, simulation results will confirm the effectiveness of the approach proposed in this paper.

  • Manabu Nii, Kazunobu Takahama, Shota Miyake, Atsuko Uchinuno, Reiko Sa ...
    原稿種別: Paper
    2014 年 18 巻 6 号 p. 918-925
    発行日: 2014/11/20
    公開日: 2019/07/01
    ジャーナル 認証あり

    Improving the quality of nursing care is crucial to maintaining the quality of life. Our objective is to develop a computer-aided evaluation system that enables nursing experts to improve the quality of nursing care. In our previous works, some classification systems based on fuzzy logic, neural networks, and SVMs were developed. Although a classification system with high performance for all nursing-care datasets is desirable, we focus on how to visualize the classification results in this paper. It is important to visualize the results for our nursing-care text classification system because the computer-aided system has to explain the reasons for obtaining such results to human experts. In this paper, a tree-type expression is considered for visualizing the classification results. To visualize classification results with the tree-type expression, we consider a decision tree technique. Word existence, dependency relations, and phrase-based feature vector definitions have been proposed in our previous works. In the present study, these three types of feature vector definitions are compared with one another from the viewpoint of understandability.

Special Issue on Forum for Interdisciplinary Mathematics 2013
  • Hiroshi Sakai, Hiroaki Ishii, Junzo Watada
    原稿種別: Editorial
    2014 年 18 巻 6 号 p. 927-928
    発行日: 2014/11/20
    公開日: 2019/07/01
    ジャーナル 認証あり

    This special issue focuses on recent research in interdisciplinary mathematics and mathematical sciences. For the last four decades, the Forum for Interdisciplinary Mathematics (FIM), a society for researchers in mathematical sciences, has focused on mathematics, combinatorics, statistics, operations research, computer science, fuzzy sets, rough sets, bioinformatics, etc.

    The 22nd International Conference of FIM on Interdisciplinary Mathematics, Statistics and Computational Techniques (IMSCT 2013-FIM XXII) was held in Kitakyushu, Japan, on November 10–12, 2013. This conference was organized by the International Society of Management (ISME international), the Graduate School of Information, Production and Systems, Waseda University, in conjunction with FIM. IMSCT 2013-FIM XXII was attended by faculty members, researchers, specialists, and graduate students from around the world.

    The 50 papers presented included keynote speeches by Professor Bhu Dev Sharma, Professor Milan Vlach, and Professor Tomonari Suzuki, together with five plenary talks. To promote FIM’s activities, guest editors had also planned to invite public participation in this special issue accepting nine papers, four selected papers from the conference and five papers closely related to this special issue. Each paper underwent strict peer reviews.

    The first paper, Crisp and Fuzzy Granular Hierarchical Structures Generated from a Free Monoid, by Tetsuya Murai, Sadaaki Miyamoto, Masahiro Inuiguchi, Yasuo Kudo, and Seiki Akama, proposes a granular hierarchy, and characterizes the mathematical structure based on fuzzy multisets, fuzzy sets, and rough multisets. This granular hierarchy includes Yager’s fuzzy multisets and Zadeh’s fuzzy sets, offering a general framework.

    The second paper, Variable Neighborhood Model for Agent Control Introducing Accessibility Relations Between Agents with Linear Temporal Logic, by Seiki Ubukata, Tetsuya Murai, Yasuo Kudo, and Seiki Akama, discusses a variable neighborhood model based on a Kripke framework, and applies this model to introducing the agent’s personal space. The authors’ research is an attempt to realize the agent’s personality.

    The third paper, Estimating Writing Neatness from Online Handwritten Data, by Motoki Miura and Takamichi Toda, considers the neatness of handwritten notes in using the authors’ AirTransNote digital pen technology. The digital pen reports physical information, and authors estimate neatness by using this physical information. Based on experiments, the authors conclude variance in pen speed, average angular point, and average pen speed are the most important features for evaluating handwriting neatness.

    The fourth paper, Application of Rough Set-Based Information Analysis to Questionnaire Data, by Naoto Yamaguchi, Mao Wu, Michinori Nakata, and Hiroshi Sakai, applies the authors’ rough non-deterministic information analysis (RNIA) to questionnaire data and question-answering. Experimental results indicate the power of the getRNIA software tool developed by the authors and possibilities for new types of data analysis.

    The fifth paper, Analysis of Consistent Equilibria in a Mixed Duopoly, by Vyacheslav V. Kalashnikov, Vladimir A. Bulavsky, Nataliya I. Kalashnykova, Junzo Watada, and Diego Je Jesús Hernández-Rodríguez, investigates a model of partially mixed duopoly with conjectured variations in equilibrium. They establish the existence and the uniqueness for conjectured variations in equilibrium for any set of feasible conjectures, and prove the existence theorem for interior equilibrium.

    The sixth paper, Mixed Oligopoly: Analysis of Consistent Equilibria, twinned with the fifth paper and by the same authors, deals with ...

    [View PDF for the rest of the abstract]

  • Tetsuya Murai, Sadaaki Miyamoto, Masahiro Inuiguchi, Yasuo Kudo, Seiki ...
    原稿種別: Paper
    2014 年 18 巻 6 号 p. 929-936
    発行日: 2014/11/20
    公開日: 2019/07/01
    ジャーナル 認証あり

    The granular hierarchical structure that we propose has the limitation that the set of truth values is not described explicitly. By regarding “sets of putting” as a special class of direct sets, we introduce crisp and fuzzy granular hierarchical structures. We found that there are two kinds of different preextensions between Yager’s fuzzy multisets and Zadeh’s fuzzy sets.

  • Seiki Ubukata, Tetsuya Murai, Yasuo Kudo, Seiki Akama
    原稿種別: Paper
    2014 年 18 巻 6 号 p. 937-945
    発行日: 2014/11/20
    公開日: 2019/07/01
    ジャーナル 認証あり

    In general, there are two types of agents, reflex and deliberative. The former does not have the ability for deep planning that produces higher-level actions to attain goals cooperatively, which is the ability of the latter. Can we cause reflex agents to act as though they could plan their actions? In this paper, we propose a variable neighborhood model for reflex agent control, that allows such agents to create plans in order to attain their goals. The model consists of three layers: (1) topological space, (2) agent space, and (3) linear temporal logic. Agents with their neighborhoods move in a topological space, such as a plane, and in a cellular space. Then, a binary relation between agents is generated each time from the agents’ position and neighborhood. We call the pair composed of a set of agents and binary relations the agent space. In order to cause reflex agents to have the ability to attain goals superficially, we consider the local properties of the binary relation between agents. For example, if two agents have a symmetrical relation at the current time, they can struggle to maintain symmetry or they could abandon symmetry at the next time, depending on the context. Then, low-level behavior, that is, the maintenance or abandonment of the local properties of binary relations, grant reflex agents a method for selecting neighborhoods for the next time. As a result, such a sequence of low-level behavior generates seemingly higher-level actions, as though reflex agents could attain a goal with such actions. This low-level behavior is shown through simulation to generate the achievement of a given goal, such as cooperation and target pursuing.

  • Motoki Miura, Takamichi Toda
    原稿種別: Paper
    2014 年 18 巻 6 号 p. 946-952
    発行日: 2014/11/20
    公開日: 2019/07/01
    ジャーナル 認証あり

    Handwriting is the most fundamental expressive activity in learning. To utilize the intuitiveness and the nature of handwriting, digital pen technology has emerged to capture and transfer notes. We developed AirTransNote, a student note-sharing system that facilitates collaborative and interactive learning in conventional classrooms. A teacher can use the AirTransNote system to share student notes with the class on a projected screen immediately to enhance the group learning experience. However, to improve the effectiveness of sharing notes, the teacher must be able to select an effective note for sharing. This can be difficult and time consuming during a lecture. Moreover, students should be encouraged to improve the presentation of their handwritten notes. Well-written notes are more accessible for other students and reduce irrelevant and careless mistakes. To facilitate learning improvements based on note sharing, we require a method to estimate the neatness of a note automatically. If a method is established, the teacher can easily select effective notes. Furthermore, this method can help provide feedback to the student to improve their writing. We examined 14 basic features from handwritten notes by considering correlation coefficients and found that the variance of pen speed, angular point average, and pen speed average were the significant features for evaluating the neatness of handwritten notes.

  • Naoto Yamaguchi, Mao Wu, Michinori Nakata, Hiroshi Sakai
    原稿種別: Paper
    2014 年 18 巻 6 号 p. 953-961
    発行日: 2014/11/20
    公開日: 2019/07/01
    ジャーナル 認証あり

    This article reports an application of Rough Nondeterministic Information Analysis (RNIA) to two data sets. One is the Mushroom data set in the UCI machine leaning repository, and the other is a student questionnaire data set. Even though these data sets include many missing values, we obtained some interesting rules by using our getRNIA software tool. This software is powered by the NIS-Apriori algorithm, and we apply rule generation and question-answering functionalities to data sets with nondeterministic values.

  • Vyacheslav V. Kalashnikov, Vladimir A. Bulavsky, Nataliya I. Kalashnyk ...
    原稿種別: Paper
    2014 年 18 巻 6 号 p. 962-970
    発行日: 2014/11/20
    公開日: 2019/07/01
    ジャーナル 認証あり

    This paper examines a model of a mixed duopoly with conjectural variations equilibrium (CVE), in which one of the agents maximizes a convex combination of his/her net profit and domestic social surplus. The agents’ conjectures concern the price variations, which depend on their production output variations. Based on the already established existence and uniqueness results for the CVE (called the exterior equilibrium) for any set of feasible conjectures, the notion of interior equilibrium is introduced by developing a consistency criterion for the conjectures (referred to as influence coefficients), and the existence theorem for the interior equilibrium (understood as a CVE state with consistent conjectures) is proven. When the convex combination coefficient tends to 1, thus transforming the model into the mixed duopoly in its extreme form, two trends are apparent. First, for the private company, the equilibrium with consistent conjectures becomes more proficient than the Cournot-Nash equilibrium. Second, there exists a (unique) value of the combination coefficient such that the private agent’s profit is the same in both of the above-mentioned equilibria, which makes subsidies to the producer or to consumers unnecessary.

  • Vyacheslav V. Kalashnikov, Vladimir A. Bulavsky, Nataliya I. Kalashnyk ...
    原稿種別: Paper
    2014 年 18 巻 6 号 p. 971-984
    発行日: 2014/11/20
    公開日: 2019/07/01
    ジャーナル 認証あり

    In this paper, a model of mixed oligopoly with conjectured variations equilibrium (CVE) is examined, in which one of the agents maximizes a convex combination of its net profit with the domestic social surplus. The agents’ conjectures concern the price variations, which depend on the variations in their production outputs. Using the established existence and uniqueness results for the CVE (the exterior equilibrium) for any fixed set of feasible conjectures, the notion of the interior equilibrium is introduced by developing a conjecture consistency criterion. Then, the existence theorem for the interior equilibrium (defined as a CVE state with consistent conjectures) is proven. When the convex combination coefficient tends to 1 (thus transforming the model into the mixed oligopoly in its extreme form), two trends are apparent. First, for private companies, the equilibrium with consistent conjectures becomes more proficient than the Cournot-Nash equilibrium. Second, there exists a (unique) value of the convex combination coefficient such that the private agent’s aggregate profit is the same in both the above-mentioned equilibria, which makes subsidies to producers or consumers unnecessary.

  • Bing Xu, Qiuqin He, Xiaowen Hu, Shangfeng Zhang
    原稿種別: Paper
    2014 年 18 巻 6 号 p. 985-991
    発行日: 2014/11/20
    公開日: 2019/07/01
    ジャーナル 認証あり

    By building a new Keynesian dynamic stochastic general equilibrium (DSGE) model, we analyze the effect of interest rate liberalization on fiscal policy. First, we find that when the interest rate increases, technology shocks, monetary policy shocks, and fiscal policy shocks can effectively stabilize economic fluctuations. Second, when the interest rate rises, fiscal policy enhances the positive effect on output first, with decreasing the negative effect on output later. Third, fiscal policy increases the original crowding-out effect on consumption and investment. However, this increase in the crowding-out effect does not restrain the positive effect of fiscal policy on output, which benefits from interest rate liberalization.

  • Toshiharu Fujita, Akifumi Kira
    原稿種別: Paper
    2014 年 18 巻 6 号 p. 992-998
    発行日: 2014/11/20
    公開日: 2019/07/01
    ジャーナル 認証あり

    In this paper, we introduce a basic framework for mutually dependent Markov decision processes (MDMDP) showing recursive mutual dependence. Our model is structured upon two types of finite-stage Markov decision processes. At each stage, the reward in one process is given by the optimal value of the alternative process problem, whose initial state is determined by the current state and decision in the original process. We formulate the MDMDP model and derive mutually dependent recursive equations by dynamic programming. Furthermore, MDMDP is illustrated in a numerical example. The model enables easier treatment of some classes of complex multi-stage decision processes.

  • Junaida Binti Sulaiman, Herdianti Darwis, Hideo Hirose
    原稿種別: Paper
    2014 年 18 巻 6 号 p. 999-1006
    発行日: 2014/11/20
    公開日: 2019/07/01
    ジャーナル 認証あり

    Successive days of precipitation are known to cause flooding in monsoon-susceptible countries. Forecasting of daily precipitation facilitates the prediction of the occurrences of rainfall and number of wet days. Using the maximum five-day accumulated precipitation (MX5d), we can predict the magnitude of precipitation in a specific period as it may indicate the extreme precipitation. In this study, a method to forecast monthly extreme precipitation using artificial neural networks (ANNs) is assessed using past MX5d data and global climate indices such as Southern Oscillation Index (SOI), Madden Julian Oscillation (MJO), and Dipole Mode Index (DMI) in Kuantan and Kota Bharu, Malaysia. The use of combined inputs (MX5d with SOI, MJO, and DMI) is proposed to investigate the concurrent effect of lagged values of local precipitation data and global climate indices on seasonal extreme precipitation. Four cases of data are sampled representing two major seasonal variations in Malaysia. The analysis of extreme precipitation trends is important for the prediction of high precipitation events. ANNs are widely applied in the hydrology field because of their nonlinear ability in predicting nonstationary and seasonal data. In this paper, we have compared ANNs with seasonal autoregressive integrated moving average (ARIMA) and regression analysis using out-of-sample test data. The results for Kuantan indicate that seasonal ARIMA is the best method to forecast extreme precipitation when MX5d lags are used as input. For Kota Bharu, ANN exhibits better generalization ability than ARIMA and regression analysis when dual inputs (lagged MX5d and lagged global climate indices) are utilized in the model.

Selected Papers from SCIS&ISIS2012
  • Keisuke Yamazaki
    原稿種別: Paper
    2014 年 18 巻 6 号 p. 1007-1012
    発行日: 2014/11/20
    公開日: 2019/07/01
    ジャーナル 認証あり

    Cluster analysis is commonly used in the fields of computational intelligence and pattern recognition. The task is to detect the unobservable labels that show to which clusters the observable data belong. A Gaussian mixture is a representative hierarchical model that is often used when taking a probabilistic approach to this task. Although it is widely used, the statistical properties of cluster analysis have not yet been clarified. The present paper analyzes the theory of Bayesian clustering for the case when the number of clusters is unknown and the variance-covariance matrix of the Gaussian distribution has a constraint. We refer to this constraint as the structure of the components. The result of this analysis shows that, even if the estimation method does not take account of the structure, the Bayes method provides an effective, tractable, and efficient algorithm. Based on an experiment with simulated data, we confirmed the advantages of the Bayes method over the expectation-maximization (EM) method.

  • Takayuki Kawaura, Yasuyuki Sugatani
    原稿種別: Paper
    2014 年 18 巻 6 号 p. 1013-1019
    発行日: 2014/11/20
    公開日: 2019/07/01
    ジャーナル 認証あり

    In the 1990s, the Japanese population aged 65 and over increased to more than 14%, and Japan became an“aging society.” Now, one in five people are 65 or over (23.4%, and one in ten people are 75 or over (1.6%, meaning that Japanese society is aging substantially. The serious problems that acute hospitals now face involve complications of diseases that are typified by deliriu, and their prevention. Patients with delirium have a higher risk of falling and dying, and delirium has a negative influence on treatment and nursing as well as on a patient’s vital prognosis. However, delirium is a mental state that is often overlooked. Thus, it is very important to develop the observation skills of staff and establish a nursing care system that does not overlook delirium. In this study, we conducted group interviews involving the clinical nurses who care for patients with delirium on a routine basis at Kansai Medical University Takii Hospital, Japan. Their spontaneous utterances about delirium were analyzed using the DEMATEL method, and these utterances were divided into two groups: “causes of delirium” and “delirious patients’ behavior.” From each group, keywords and phrases were chosen and analyzed. Consequently, this study will demonstrate how these clinical nurses feel about delirium and delirious patients.

  • Maki Sakamoto, Yuya Ueda, Ryuichi Doizaki, Yuichiro Shimizu
    原稿種別: Paper
    2014 年 18 巻 6 号 p. 1020-1025
    発行日: 2014/11/20
    公開日: 2019/07/01
    ジャーナル 認証あり

    In medical interviews, Japanese patients often use onomatopoeia, such as ‘zuki-zuki’ and ‘chiku-chiku,’ to express pain symptoms and medical conditions. However, onomatopoeia shows cross-linguistic variation, and thus Japanese onomatopoeia cannot be used effectively to express pain symptoms in medical interviews with foreign doctors who do not speak Japanese. In this study, we developed a system that supports communication between Japanese patients and foreign doctors by putting an onomatopoeia evaluation system to medical use. Our system estimates the quality of pain and other medical conditions based on the sound symbolic meanings expressed by certain onomatopoetic expressions. The relationships between the sound symbolic properties and rating scales were obtained through psychological experiments in which 120 participants evaluated the mental images of 354 Japanese onomatopoeia terms used to express pain symptoms and medical conditions against 35 semantic differential (SD) scales such as “sharp–dull,” “strong–weak,” and “momentary–continuous.” Our system accepts any Japanese onomatopoetic expression input by users and can also respond to any novel onomatopoetic expression. If the rating scales were translated into various languages, foreign doctors all over the world would be able to understand the meaning of Japanese onomatopoeia.

  • Okito Yamashita, Takeaki Shimokawa, Takashi Kosaka, Takashi Amita, Yos ...
    原稿種別: Paper
    2014 年 18 巻 6 号 p. 1026-1033
    発行日: 2014/11/20
    公開日: 2019/07/01
    ジャーナル 認証あり

    Diffuse optical tomography (DOT) is an emerging technology for improving the spatial resolution of conventional multi-channel near infrared spectroscopy (NIRS). The hemodynamics changes in two distinct anatomical layers, the scalp and the cortex, are known as the main contributor of NIRS measurement. Although any DOT algorithm has the ability to reconstruct scalp and cortical hemodynamics changes in their respective layers, no DOT algorithm has used a model characterizing the distinct nature of scalp and cortical hemodynamics changes to achieve accurate separation. Previously, we have proposed a hierarchical Bayesian model for DOT in which distinct prior distributions for the scalp and the cortical hemodynamics changes are assumed and then verified the reconstruction performance with a phantom experiment and a computer simulation of a real human head model (Shimokawa et al. 2013, Biomedical Optical Express). Here, we investigate the reconstruction accuracy of the proposed algorithm using human experimental data for the first time. We measured the brain activities of a single subject during a finger extension task with NIRS and fMRI. Our DOT reconstruction was compared with the fMRI localization results. Consequently, a remarkable consistency between fMRI and our DOT reconstruction was observed both in the spatial and temporal patterns. By extending the advantages of NIRS such as low running cost and portability with our DOT method, it might be possible to advance brain research in a real environment, which cannot be done with fMRI.

  • Keisuke Kimura, Takayuki Kimura, Takefumi Hiraguri, Kenya Jin’no
    原稿種別: Paper
    2014 年 18 巻 6 号 p. 1034-1043
    発行日: 2014/11/20
    公開日: 2019/07/01
    ジャーナル 認証あり

    In this paper, the control method based on recurrent neural networks is proposed for optimizing large-scale wind and solar power generation systems. Recently, an optimal control method based on recurrent neural networks was proposed for wind and solar power generation systems. In this method, optimization problems are regarded as linear programming problems, which are solved by recurrent neural networks. Results suggest that this control method based on recurrent neural networks could be implemented in real-world systems. However, only small power generation systems were used to evaluate this control method in previous studies. Then, the method for power generation systems is evaluated by more realistic conditions. The results of our numerical experiments show that this control method delivers high performance with large-scale power generation systems. Furthermore, if the power generation systems has specific topologies, almost 20% of the supplying capacity is improved.

  • Takaki Urai, Masataka Tokumaru
    原稿種別: Paper
    2014 年 18 巻 6 号 p. 1044-1052
    発行日: 2014/11/20
    公開日: 2019/07/01
    ジャーナル 認証あり

    This study proposes a clothing-image retrieval system that considers a user’s kansei, i.e., the user’s emotional, physical, and aesthetical preferences. The system gradually learns the user’s kansei by repeated interaction with the user, and can then search for corresponding clothing images. The practicality of the proposed system is confirmed in experiments with real users. It is found that the proposed system can retrieve clothing images that corresponded to the user’s kansei.

feedback
Top