Journal of Japan Society for Fuzzy Theory and Intelligent Informatics
Online ISSN : 1881-7203
Print ISSN : 1347-7986
ISSN-L : 1347-7986
Volume 30 , Issue 4
Showing 1-10 articles out of 10 articles from the selected issue
Special Feature: Selected paper from Heart and Mind 2017
Original Papers
  • Makoto OHKI, Kaito KUDOU, Genki TOKUNAGA
    2018 Volume 30 Issue 4 Pages 605-612
    Published: August 15, 2018
    Released: August 15, 2018
    JOURNALS OPEN ACCESS

    This paper proposes to measure of the Team Orientation (consensus degree of a purpose) in a group decision making on a business. A group demonstrates big capability if a purpose of the group is agreed from the members. If we have the measure of the Team Orientation in a group decision making, then it contributes to activation and increase in efficiency of organizations greatly. The bosses and leaders can measure working efficiency using the Team Orientation. I proposed the “distance-adjusted covariates method” for group decision making analysis. This method can categorize group to 4 areas using “VDI (Variety Dispersion Index)” and “ranking values”. These are defined in this method. In this paper, I show that the VDI expresses the Team Orientation in a group decision making. A business simulation game is used as a practical scene of an experiment in this paper. It can make a practical scene of a business. This paper aims to show that the proposed method can analyze in detail a group’s behavior on practical scene of business using VDI. As a result, the signs that the group’s purpose going in agreement were evaluated by the VDI. And we found that if a team has high consensus degree then the team has high performance.

    Download PDF (1231K)
  • Minatsu FUJISAKI, Hiroshi TAKENOUCHI, Masataka TOKUMARU
    2018 Volume 30 Issue 4 Pages 613-622
    Published: August 15, 2018
    Released: August 15, 2018
    JOURNALS OPEN ACCESS

    We propose an interactive evolutionary computation(IEC)system using gaze information from multiple users. IEC is an optimization method that can reflect human sensibility in the generation of solution candidates. IEC studies have been applied to a wide range of fields. However, the drawback of IEC is that the user evaluation load is large. To solve this problem, we propose using the user’s gaze information in the IEC evaluation. The proposed system acquires gaze information when a user is viewing a solution candidate. From this gaze information, the system assigns evaluation values to the solution candidate. Then it is unnecessary for the user to give an evaluation of the solution candidate. Therefore, the proposed system effectively reduces the user’s evaluation load. In addition, it is possible to obtain multiple gaze information at the same time. Therefore, in the proposed method, it is possible to simultaneously reflect the sensitivity of multiple users to the IEC solution candidate evaluation. We developed a female clothes coordination generation system using the proposed method and verified the evolution performance of the solution candidate via the proposed method and the effect on the users’ evaluation loads. The proposed system was able to generate satisfactory solution candidates to some extent without feeling heavy evaluation loads. Therefore, we confirmed the effectiveness of the proposed method. In addition, we confirmed the differences in reactions between men and women during the experiment, as well as the differences in the appearance rate of clothing coordination. As a result, we confirmed the difference in the evaluation method for clothes coordination in men and women.

    Download PDF (5136K)
Regular
Short Notes
  • Reona GOMI, Shinya IWASAKI, Eri SHIMOKAWARA, Toru YAMAGUCHI
    2018 Volume 30 Issue 4 Pages 623-627
    Published: August 15, 2018
    Released: August 15, 2018
    JOURNALS OPEN ACCESS

    A telepresence robot is one of the remote conversation support robots. As an example for a main transmission movement of telepresence robot includes head movement. There are many previous researches dealing with convection of head movement by telepresence robot. However, these researches don’t deal with “screen movement” by head movement. In this paper, the authors indicate the impact of screen movement on users in conversation and to develop a telepresence robot that transmits only the listener’s nod to decrease screen movement. Moreover, we indicate usefulness of developed telepresence robot by comparison experiment between developed telepresence robot and a robot transmitting many screen movements.

    Download PDF (921K)
  • Ryota NISHIMURA, Miho HIGAKI, Norihide KITAOKA
    2018 Volume 30 Issue 4 Pages 628-633
    Published: August 15, 2018
    Released: August 15, 2018
    JOURNALS OPEN ACCESS

    In this research, we propose a method to search between different media(cross media mapping)using Recurrent Neural Network(RNN)which is a kind of Neural Network(NN). By using the proposed method, it becomes possible to correlate music and lyrics, it is possible to search music using documents. Application of this model realizes a system that provides appropriate BGM from conversation contents by monitoring human interactions. In this paper, we constructed a proposal model, conducted an evaluation experiment, and confirmed the possibility of cross media mapping.

    Download PDF (1423K)
Original Papers
  • Komei HASEGAWA, Seigo FURUYA, Yusuke KANAI, Kazuhiko SHINOZAWA, Michit ...
    2018 Volume 30 Issue 4 Pages 634-642
    Published: August 15, 2018
    Released: August 15, 2018
    JOURNALS OPEN ACCESS

    This paper proposes DECoReS(Degree Expressional Command Reproducing System)that allows an autonomous wheelchair to travel autonomously through degree expressional commands. When users control an autonomous wheelchair through voice commands, they can sometimes give such orders as “go straight speedily” and “make a wide right turn.” These optional words called “degree expression” are appended to the commands to adapt the movement to the user’s preference. Because degree expressions are ambiguous, traveling styles described with degree expressions are altered depending on the users and environments. DECoReS realizes the movement adapted for each user by learning degree expressional commands, traveling data from the users, and map data. DECoReS also reproduces travels suited for users and his/her degree expressions even for a different environment by extracting a similar map from learned data. Our experiments show that DECoReS can reproduce different travels depending on degree expressional commands, individual users, and some different environments.

    Download PDF (2839K)
  • Kazuma KASAHARA, Kenta NIMOTO, Takashi ITO, Kenichi TAKAHASHI, Michima ...
    2018 Volume 30 Issue 4 Pages 643-651
    Published: August 15, 2018
    Released: August 15, 2018
    JOURNALS OPEN ACCESS

    This paper proposes a method to improve the learning performance of a learning agent with FALCON, to make a player agent for the card game “Hearts”, which is one of the multi-player imperfect-information games. FALCON is a machine learning method which is an extended fuzzy ART(Adaptive Resonance Theory). The previous work showed that FALCON is effective for Hearts. In this study, to improve the learning performance, the action set of the agent is changed based on strategies of Hearts, and a method that employs a prediction by the support vector regression is proposed.

    Download PDF (1189K)
  • Reo HOSHINO, Takayuki SHIINA, Susumu MORITO, Jun IMAIZUMI
    2018 Volume 30 Issue 4 Pages 652-657
    Published: August 15, 2018
    Released: August 15, 2018
    JOURNALS OPEN ACCESS

    In this research, we aim to create a schedule in professional baseball that makes the difference of remaining game number among teams as small as possible. We consider not only making of deterministic schedule, but also a schedule that takes into account uncertainty such as cancellation due to rain. First, a schedule is generated by integer programming. By considering the rainout, we try to minimize the difference in the number of remaining games among the teams. In this research, we reduced the number of remaining games compared with the actual schedule, the difference between the maximum expected value and the minimum value of each team could also be suppressed. As a result, it is possible to reduce difference in the number of remaining games among the teams.

    Download PDF (711K)
  • Tomohiro HAYASHIDA, Ichiro NISHIZAKI, Shinya SEKIZAKI, Hiroaki TAKEUCH ...
    2018 Volume 30 Issue 4 Pages 658-665
    Published: August 15, 2018
    Released: August 15, 2018
    JOURNALS OPEN ACCESS

    A classifier sysem ACSM (ACS 2 with Memory) (Hayashida et al., 2014) is developed on the basis of ACS (Anticipatory Classifier System) (Stolzmann, 1997, 1998) and ACS2 (Butz and Stolzmann, 2002) which is an improved system of ACS. ACSM includes internal memory to distinguish the aliased states of POMDPs (Partially Observable Markov Decision Process) and avoid some wasteful searching process. This paper optimize the complicated procedure of the learning process of ACSM, and aims to reduce unnecessary computation time or improve learning performance. Numerical experiments are executed using some maze problems which are adopted in a number of related papers as benchmark problems of POMDPs including aliased states. The experimental results indicate the proposed method can find the solutions more efficiently compared to ACS2 and ACSM.

    Download PDF (1117K)
feedback
Top