Researchers Database

Yoshida Tetsuya

    Faculty Division of Engineering Research Group of Engineering Professor
Last Updated :2021/06/02

researchmap

Degree

  • (BLANK), The University of Tokyo
  • (BLANK), University of Edinburgh

Research Interests

  • 機械学習 人工知能 知能情報学 

Research Areas

  • Informatics, Intelligent informatics

Research Experience

  • Apr. 2014, Nara Women's University, 教授
  • Apr. 2004 Mar. - 2014, Hokkaido University, Graduate School of Information Science and Technology, Division of Computer Science, 准教授
  • Oct. 2001 Mar. - 2004, Osaka University, The Institute of Scientific and Industrial Research, 助手

Education

  • Apr. 1994, Mar. - 1997, The University of Tokyo, 工学系研究科, 先端学際工学専攻, Japan
  • Apr. 1991, Mar. - 1994, The University of Tokyo, The Graduate School of Engineering, 航空宇宙工学専攻
  • Oct. 1992, Sep. - 1993, University of Edinburgh, Department of Artificial Intelligence, 知識処理専攻
  • Apr. 1987, Mar. - 1991, The University of Tokyo, The Faculty of Engineering, 航空学科, Japan

Published Papers

  • 正多角形の貼り合わせを用いた花紋スモッキングの 組み合わせの拡張

    夛田美沙; 吉田哲也

    Nov. 2020, 芸術科学会論文誌, 19 (4), 40 - 48

  • 花紋折りに基づくスモッキングのパターン作成と組 み合わせのデザイン

    吉田哲也,藤﨑千晶

    Jun. 2020, 芸術科学会論文誌, 19 (2), 9 - 24

    Scientific journal

  • ネパー ルの世界文化遺産登録都市における都市型住居の外観意匠類型 – バクタプ ル東部の都市街区を事例に–

    飛鳥濱岡; 山本直彦; 吉田哲也; 宮内杏里; 増井正哉; 向井洋一

    Jan. 2019, 日本建築学会計画系論文集, 84 (756), 425 - 435

    Scientific journal

  • COMMUNITY STRUCTURE-BASED APPROACH FOR NETWORK IMMUNIZATION

    Tetsuya Yoshida; Yuu Yamada

    We propose a community structure-based approach that does not require community labels of nodes, for network immunization. Social networks have been widely used as daily communication infrastructures these days. However, fast spreading of information over networks may have downsides such as computer viruses or epidemics of diseases. Because contamination is propagated among subgraphs (communities) along links in a network, use of community structure of the network would be effective for network immunization. However, despite various research efforts, it is still difficult to identify ground-truth community labels of nodes in a network. Because communities are often interwoven through intermediate nodes, we propose to identify such nodes based on the community structure of a network without requiring community labels. By regarding the community structure in terms of nodes, we construct a vector representation of nodes based on a quality measure of communities. The distribution of the constructed vectors is used for immunizing intermediate nodes among communities, through the hybrid use of the norm and the relation in the vector representation. Experiments are conducted over both synthetic and real-world networks, and our approach is compared with other network centrality-based approaches. The results are encouraging and indicate that it is worth pursuing this path., WILEY, Feb. 2017, COMPUTATIONAL INTELLIGENCE, 33 (1), 77 - 98, doi;web_of_science

    Scientific journal

  • A Graph-based approach for Semi-Supervised Clustering

    Yoshida, T

    May 2014, Computational Intelligence, 30 (2), 263 - 284

    Scientific journal

  • Weighted line graphs for overlapping community discovery

    Tetsuya Yoshida

    Springer Science and Business Media LLC, Dec. 2013, Social Network Analysis and Mining, 3 (4), 1001 - 1013, doi;url;url;url

    Scientific journal

  • Rectifying the representation learned by Non-negative Matrix Factorization

    Yoshida, T

    Nov. 2013, International Journal of Knowledge-Based & Intelligent Engineering Systems, 17 (4), 279 - 290

    Scientific journal

  • Toward finding hidden communities based on user profile

    Tetsuya Yoshida

    Springer Science and Business Media LLC, Apr. 2013, Journal of Intelligent Information Systems, 40 (2), 189 - 209, doi;url;url;url

    Scientific journal

  • Toward Robust and Fast Two-Dimensional Linear Discriminant Analysis

    Tetsuya Yoshida; Yuu Yamada

    This paper presents an approach toward robust and fast Two-Dimensional Linear Discriminant Analysis (2DLDA). 2DLDA is an extension of Linear Discriminant Analysis (LDA) for 2-dimensional objects such as images. Linear transformation matrices are iteratively calculated based on the eigenvectors of asymmetric matrices in 2DLDA. However, repeated calculation of eigenvectors of asymmetric matrices may lead to unstable performance. We propose to use simultaneous diagonalization of scatter matrices so that eigenvectors can be stably calculated. Furthermore, for fast calculation, we propose to use approximate decomposition of a scatter matrix based on its several leading eigenvectors. Preliminary experiments are conducted to investigate the effectiveness of our approach. Results are encouraging, and indicate that our approach can achieve comparative performance with the original 2DLDA with reduced computation time., SPRINGER-VERLAG BERLIN, 2013, ACTIVE MEDIA TECHNOLOGY, AMT 2013, 8210, 126 - 135, web_of_science

    International conference proceedings

  • 重複コミュニティ発見のための重み付き線グラフ

    吉田 哲也

    Jul. 2012, 情報処理学会論文誌:数理モデル化と応用,, 5 (3), 79 - 88

    Scientific journal

  • ネットワークのノード情報を考慮した正則化モジュラリティ固有空 間法

    吉田 哲也

    Jul. 2012, 情報処理学会論文誌:数理モデル化と応用,, 5 (1), 65 - 73

    Scientific journal

  • Immunization of Networks via Modularity Based Node Representation

    Tetsuya Yoshida; Yuu Yamada

    We propose an approach for immunization of networks via modularity based node representation. Immunization of networks has often been conducted by removing nodes with large centrality so that the whole network can be fragmented into smaller subgraphs. Since contamination is propagated among subgraphs (communities) along links in a network, besides centrality, utilization of community structure seems effective for immunization. However, despite various efforts, it is still difficult to identify true community labels in a network. Toward effective immunization of networks, we propose to remove nodes between communities without identifying community labels of nodes. By exploiting the vector representation of nodes based on the modularity matrix of a network, we propose to utilize not only the norm of vectors, but also the relation among vectors. Two heuristic scoring functions are proposed based on the inner products of vector representation and their filtering in terms of vector angle. Preliminary experiments are conducted over synthetic networks and real-world networks, and compared with other centrality based immunization strategies., SPRINGER-VERLAG BERLIN, 2012, INTELLIGENT DECISION TECHNOLOGIES (IDT'2012), VOL 2, 16, 33 - 44, doi;web_of_science

    International conference proceedings

  • Line Graph for Weighted Networks toward Overlapping Community Discovery

    Tetsuya Yoshida

    We propose generalized line graph for weighted networks toward overlapping community discovery from the networks. Community discovery from a network has often been conducted by assigning each node in a network only to one community. However, in real world networks, a node (e.g., user) might belong to several communities. For undirected networks without self-loops, we propose to generalize line graph by defining the weights in the line graph based on the weights in the original network. Based on the line graph representation, a node can be assigned to more than one community by assigning the links adjacent to the node to the corresponding communities. Various properties of the proposed generalized line graph are clarified, and the properties indicate that our proposal is a natural extension of the conventional line graph. Preliminary experiments are conducted over several real-world networks, and the results indicate that the proposed generalized line graph can improve the quality of the discovered overlapping communities., SPRINGER-VERLAG BERLIN, 2012, INTELLIGENT DECISION TECHNOLOGIES (IDT'2012), VOL 2, 16, 403 - 413, web_of_science

    International conference proceedings

  • A comparative study of community structure based node scores for network immunization

    Yuu Yamada; Tetsuya Yoshida

    Network immunization has often been conducted by removing nodes with large network centrality so that the whole network can be fragmented into smaller subgraphs. Since contamination (e.g., virus) is propagated among subgraphs (communities) along links in a network, besides centrality, utilization of community structure seems effective for immunization. We have proposed community structure based node scores in terms of a vector representation of nodes in a network. In this paper we report a comparative study of our node scores over both synthetic and real-world networks. The characteristics of the node scores are clarified through the visualization of networks. Extensive experiments are conducted to compare the node scores with other centrality based immunization strategies. The results are encouraging and indicate that the node scores are promising. © 2012 Springer-Verlag., 2012, Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), 7669, 328 - 337, doi

    International conference proceedings

  • Influence of erroneous pairwise constraints in semi-supervised clustering

    Tetsuya Yoshida

    Side information such as pairwise constraints is useful to improve the clustering performance in general. However, constraints are not always error free in general. When erroneous constraints are specified as side information, treating them as hard constraints could have the disadvantage since strengthening incorrect or erroneous constraints can lead to performance degradation. In this paper we conduct extensive experiments to investigate the influence of erroneous pairwise constraints over various document datasets. Several state-of-the-art semi-supervised clustering methods with graph representation were evaluated with respect to the type of constraints as well as the number of constraints. Experimental results confirmed that treating pairwise constraints as hard constraints is vulnerable to erroneous ones. However, the results also revealed that the influence of erroneous constraints depends on how the constraints are exploited inside a learning algorithm. © 2012 Springer-Verlag., 2012, Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), 7669, 43 - 52, doi

    International conference proceedings

  • Community structure based node scores for network immunization

    Tetsuya Yoshida; Yuu Yamada

    We propose community structure based node scores for network immunization. Since epidemics (e.g, virus) are propagated among groups of nodes (communities) in a network, network immunization has often been conducted by removing nodes with large score (e.g., centrality) so that the major part of the network can be protected from the contamination. Since communities are often interwoven through intermediating nodes, we propose to identify such nodes based on the community structure of a network. By regarding the community structure in terms of nodes, we construct a vector representation of each node based on a quality measure of communities for node partitioning. Two types of node score are proposed based on the direction and the norm of the constructed node vectors. © 2012 Springer-Verlag., 2012, Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), 7458, 899 - 902, doi

    International conference proceedings

  • Overlapping community discovery via weighted line graphs of networks

    Tetsuya Yoshida

    We propose an approach for overlapping community discovery via weighted line graphs of networks. For undirected connected networks without self-loops, we generalize previous weighted line graphs by: 1) defining weights of a line graph based on the weights in the original network, and 2) removing self-loops in weighted line graphs, while sustaining their properties. By applying some off-the-shelf node partitioning method to the weighted line graph, a node in the original network can be assigned to more than one community based on the community labels of its adjacent links. Various properties of the proposed weighted line graphs are clarified. Furthermore, we propose a generalized quality measure for soft assignment of nodes in overlapping communities. © 2012 Springer-Verlag., 2012, Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), 7458, 895 - 898, doi

    International conference proceedings

  • A re-coloring approach for graph b-coloring based clustering

    Tetsuya Yoshida; Hiroki Ogino

    This paper proposes a re-coloring approach for graph b-coloring based clustering. Based on the notion of graph b-coloring in graph theory, a b-coloring based clustering method was proposed. However, previous method did not explicitly consider the quality of clusters, and could not find out better clusters which satisfy the properties of b-coloring. Although a greedy re-coloring algorithm was proposed to reflect the quality of clusters, it was still restrictive in terms of the explored search space due to its greedy and sequential re-coloring process. We aim at overcoming the limitations by enlarging the search space for re-coloring, while guaranteeing the b-coloring properties. In our approach, the vertices in a graph are divided into two disjoint subsets based on the properties of b-coloring. A best first re-coloring algorithm is proposed to realize non-greedy search for the admissible colors of vertices. A color exchange algorithm is proposed to remedy the problem in sequential re-coloring. These algorithms are orthogonal to each other with respect to the re-colored vertices, and thus can be utilized in conjunction. The proposed approach was evaluated against several UCI benchmark datasets. The results are encouraging and indicate the effectiveness of the proposed method, especially with respect to the ground truth micro-averaged precision. © 2012 - IOS Press and the authors. All rights reserved., 2012, International Journal of Knowledge-Based and Intelligent Engineering Systems, 16 (2), 117 - 128, doi

    Scientific journal

  • 特徴表現のスパース性を考慮したNMF

    木村 圭吾; 吉田 哲也

    2012, 情報処理学 会論文誌:数理モデル化と応用, 5 (1), 21 - 29

    Scientific journal

  • A graph model for mutual information based clustering

    Tetsuya Yoshida

    Springer Science and Business Media LLC, Oct. 2011, Journal of Intelligent Information Systems, 37 (2), 187 - 216, doi;url;url;url

    Scientific journal

  • NMFを用いた表現学習に対するコレスキー分解を用い た補正法

    吉田哲也; 荻野広樹

    Jul. 2011, 4 (3), 94 - 101

  • Fast apriori-based graph mining algorithm and application to 3-dimensional structure analysis

    Yoshio Nishimura; Takashi Washio; Tetsuya Yoshida; Hiroshi Motoda; Akihiro Inokuchi; Takashi Okada

    Apriori-based Graph Mining (AGM) algorithm efficiently extracts all the subgraph patterns which frequently appear in graph structured data. The algorithm can deal with general graph structured data with multiple labels of vartices and edges, and is capable of analyzing the topological structure of graphs. In this paper, we propose a new method to analyze graph structured data for a 3-dimensional coordinate by AGM. In this method the distance between each vertex of a graph is calculated and added to the edge label so that AGM can handle 3-dimensional graph structured data. One problem in our approach is that the number of edge labels increases, which results in the increase of computational time to extract subgraph patterns. To alleviate this problem, we also propose a faster algorithm of AGM by adding an extra constraint to reduce the number of generated candidates for seeking frequent subgraphs. Chemical compounds with dopamine antagonist in MDDR database were analyzed by AGM to characterize their 3-dimensional chemical structure and correlation with physiological activity., 2003, Transactions of the Japanese Society for Artificial Intelligence, 18 (5), 257 - 268, doi

    Scientific journal

  • A Method for Detecting Conceptual Difference based on Diverse Structure

    Teruyuki Kondo; Shogo Nishida; Tetsuya Yoshida

    When people carry out collaborative works, conceptual difference due to different backgrounds and knowledge can hinder their communication and deteriorate collaboration. We have been carrying out research on detecting conceptual difference by focusing on the situation in which different symbols are used to denote the same meaning and the same symbols are used to denote different meaning. In our approach each user's knowledge is represented as a decision tree respectively and difference in concept is detected as difference in the structure of decision trees. This paper points out some problems in constructing a unique decision tree based on a single information criterion. Based on the idea of diverse structure, this paper proposes a new method for increasing the performance of detection by constructing multiple decision trees with diverse structure. Genetic algorithm is utilized to realize the idea of diverse structure and experiments with motor diagnosis cases on the implemented system confirmed the i provement on the performance of detection. © 2003, The Institute of Electrical Engineers of Japan. All rights reserved., 2003, IEEJ Transactions on Electronics, Information and Systems, 123 (2), 345 - 354, doi

    Scientific journal

  • A Method for Detecting Conceptual Difference Based on Correlation between Decision Trees

    OHNISHI Kensuke; YOSHIDA Tetsuya; NISHIDA Shogo

    複数の人間が共同作業を行う場合には,背景知識や専門領域の違いから派生する概念の相違が意思疎通を阻害する要因となる.筆者らは従来から概念の相違として異なるシンボルを同じ意味に用いたり,同じシンボルを異なる意味に用いたりする場合を考え,ユーザの知識を決定木として表現し,概念の相違を決定木の構造の相違としてとらえることでその検出を行うことに取り組んできた.本論文では,従来の各ユーザの知識から仮想的な知識を生成して決定木を構築することにより相違を検出する手法における問題点を指摘し,その解決策として決定木におけるクラス分類の行われ方からボトムアップに決定木間の相関関係を構築して相違検出を行う手法を提案する.提案した手法をプロトタイプに実装し,モータの故障診断事例に対する実験を通じて検出精度の向上を確認した., The Institute of Electronics, Information and Communication Engineers, 25 Jul. 2002, The transactions of the Institute of Electronics, Information and Communication Engineers. D-I, 85 (8), 784 - 797, cinii_articles

  • 決定木の相関関係に基づいた概念相違検出手法

    2002, 電子情報通信学会論文誌D-I, vol.J85, No.8, pp.784-797

    Scientific journal

  • An Information Acquisition Support System based on Keyword Relationship

    SHINKAI Daiki; YOSHIDA Tetsuya; NISHIDA Shogo

    2002, Human interface. The Transaction of Human Interface Society, Vol.4, No.4,pp.19-30, cinii_articles

    Scientific journal

  • An Image Based Support System for Web Page Design

    WATANABE Masato; YOSHIDA Tetsuya; SAIWAKI Naoki; HIJIKATA Yoshinori; NISHIDA Shogo

    2001, Human interface. The Transaction of Human Interface Society, Vol.3, No.4,pp.287-297, cinii_articles

    Scientific journal

  • Utilizing the correlation between query keywords for information retrieval

    T Yoshida; D Shinkai; S Nishida

    This paper proposes a method to utilize the correlation between query keywords in information retrieval toward extrapolating their semantic information. In accordance with the rapid development of internet and WWW (World Wide Web), it has been getting more and more hard to pinpoint the appropriate document from the huge information resource. Various search engines have been developed to retrieve the appropriate information based on the keywords, however, it is hard for the user to specify the keywords enough to pinpoint the appropriate documents. Since there often exist some semantic correlation between the keywords, this paper proposes to finding out the another keyword by utilizing their semantic correlation in order to narrow the scope of seach. Experiments were carried out to investigate the effectiveness of the proposed method and the result hinted the effectiveness of our approach as a pre-processing to narrow the scope of search for search engines., SPRINGER-VERLAG BERLIN, 2001, ADVANCES IN WEB-AGE INFORMATION MANAGEMENT, PROCEEDINGS, 2118, 49 - 59, web_of_science

    Scientific journal

  • Adaptive Hypermedia System for Supporting Information Providers to Direct Users through Hyperspace

    HIJIKATA Yoshinori; YOSHIDA Tetsuya; NISHIDA Shogo

    電気学会, 01 Nov. 2000, The Transactions of the Institute of Electrical Engineers of Japan. C, 120 (11), 1720 - 1731, cinii_articles

  • ユーザ参加型設計のそれの事例を利用した意図理解支援インタフェース

    2000, ヒューマンインタフェース学会論文誌, 2 (2), 87 - 95

    Scientific journal

  • A method for supporting software design based on comment on software

    T Yoshida; K Hashimoto; T Yamaoka; S Nishida

    This paper proposes a method for supporting software design by utilizing the comment which is usually put on software. It is believed that the software designer often leaves the hint or clue with respect to the reason or rationale for each module or component as the comment on software when he/she designs and implements the software. The comment is utilized for structuring: software into a tree and CBR (Case-Based Reasoning) is carried out based on the tree structure for enabling; the effective re-use of past software with respect to the structure and organization of software in the uppser stream in design. A prototype system has been implemented with the proposed support method and experiments have been carried out to investigate the effectiveness of our approach., IEEE, 2000, IEEE RO-MAN 2000: 9TH IEEE INTERNATIONAL WORKSHOP ON ROBOT AND HUMAN INTERACTIVE COMMUNICATION, PROCEEDINGS, 311 - 315, web_of_science

    International conference proceedings

  • A Communication Model in Emergency Which Considers Competence, Duty, Responsibility and Knowledge

    KOISO Takashi; YOSHIDA Tetsuya; SAIWAKI Naoki; NISHIDA Shogo

    大規模災害に対処する防災システムを考えるとき, 適切な部署に,適切な情報が適切なタイミングで届くコミュニケーシヨンの実現が重要である. 本論文では, 階層型組織での緊急時の意思決定について分析するとともに, 組織の各構成員のもつ権限・義務・責任・知識に注目することにより, 人的構造を考慮した緊急時のコミュニケーションモデルを提案する. 更にこのコミュニケーションモデルのプラント制御への適用例を示すとともに, このモデルの適用可能性として緊急時のコミユニケーシヨン支援システムへの応用や, 組織形態のコミュニケーシヨンの視点から見た評価についても論じる., 一般社団法人電子情報通信学会, Mar. 1999, The Transactions of the Institute of Electronics, Information and Communication Engineers. A, 82 (3), 445 - 453, cinii_articles

    Scientific journal

  • Evolving granules for classification for discovering difference in the usage of words

    T Yoshida; T Kondo; S Nishida

    This paper proposes an evolutionary approach for discovering difference in the usage of words to facilitate collaboration among people. When people try to communicate their concepts with words, the difference in the meaning and usage of words can lead to misunderstanding in communication, which can hinder their collaboration. In our approach each granule of knowledge in classification from users is structured into a decision tree so that difference in the usage of word can be discovered as the difference in the structure of tree. By treating each granule of classification knowledge as an individual in Genetic Algorithm (GA), evolution is carried out with respect to the classification efficiency of each individual and diversity as a population so that difference in the usage of words will emerge as the difference in the structure of decision tree. Experiments were carried out on motor diagnosis cases with artificially encoded difference in the usage of words and the result shows the effectiveness of our evolutionary approach., SPRINGER-VERLAG BERLIN, 1999, NEW DIRECTIONS IN ROUGH SETS, DATA MINING, AND GRANULAR-SOFT COMPUTING, 1711, 366 - 374, web_of_science

    Scientific journal

  • Learning the Balance between Exploration and Exploitation via Reward

    1999, IEICE Transactions on Fundamentals of Electronics, Communications and Computer Sciences, 82-EA (11), 2538 - 2545

    Scientific journal

  • Design of the Interfaces to Detect Conceptual Difference among Different People(on Next Generation Human Interface and Interaction)

    KONDO TERUYUKI; YOSHIDA TETSUYA; NISHIDA SHOGO

    Conceptual difference among different people is a serious problem when people work in collabolation with others. In this paper, we study about conceptual difference among different people and describe a method of detecting conceptual difference by the decision trees constructed by users' knowledge in the cases that the same symbols are used in the different meaning and the cases that the different symbols are used in the same meaning. And a prototype sysytem is evaluated., 一般社団法人情報処理学会, May 1998, IPSJ Journal, 39 (5), 1195 - 1202, cinii_articles

    Scientific journal

  • A cooperation method via metaphor of explanation

    T Yoshida; K Hori; S Nakasuka

    This paper proposes a new method to improve cooperation in concurrent systems within the framework of Multi-Agent Systems (MAS). Since subsystems work concurrently, achieving appropriate cooperation among them is important to improve the effectiveness of the overall system. When subsystems are modeled as agents, it is easy to explicitly deal with the interactions among them since they can be modeled naturally as communication among agents with intended information. Contrary to previous approaches which provided the syntax of communication protocols without semantics, we focus on the semantics of cooperation in MAS and aim at allowing agents to exploit the communicated information for cooperation. This is attempted by utilizing more coarse-grained communication based on the different perspective for the balance between Formality and richness of communication contents so that each piece of communication contents can convey more meaningful information in application domains. In our approach agents cooperate each other by giving Feedbacks based on the metaphor of explanation which is widely used in human interactions, in contrast to previous approaches which use direct orders given by the leader based on the pre-defined cooperation strategies. Agents show the difference between the proposal and counterproposals for it, which are constructed with respect to the Former and given as the feedbacks in the easily understandable terms For the receiver. From the comparison of proposals agents retrieve the information on which parts are agreed and disagreed by the relevant agents, and reflect the analysis in their following behavior. Furthermore, communication contents are annotated by agents to indicate the degree of importance in decision making for them, which contributes to making explanations or feedbacks more understandable. Our cooperation method was examined through experiments on the design of micro satellites and the result showed that it was effective to some extent to facilitate cooperation among agents., IEICE-INST ELECTRONICS INFORMATION COMMUNICATIONS ENG, Apr. 1998, IEICE TRANSACTIONS ON FUNDAMENTALS OF ELECTRONICS COMMUNICATIONS AND COMPUTER SCIENCES, E81A (4), 576 - 585, web_of_science

    Scientific journal

  • 複数の人間における概念相違検出のためのインターフェースの設計

    1998, 情報処理学会論文誌, 39 (5), 1195 - 1202

    Scientific journal

  • A reinforcement learning approach to cooperative problem solving

    T Yoshida; K Hori; S Nakasuka

    We propose an extension of reinforcement learning methods to cooperative problem solving in multi-agent systems. Exploiting multiple agents for complex problems is promising, however, learning is necessary since complete domain knowledge is rarely available. The temporal difference algorithm is applied in each agent to learn a heuristic evaluation of states. Besides the reward for solutions produced by agents, we define the reward for coherence as a. whole and exploit them to facilitate cooperation among agents for global problem solving. We evaluate the method by experiments on the satellite design problem. The result shows that our method enables agents to learn to cooperate as well as to learn individual heuristics within one framework. Especially, agents themselves learn to take the appropriate balance between exploration and exploitation in problem solving, which is known to greatly affect the performance. It also suggests the possibility of controlling the global behavior of multi-agent systems via rewards in reinforcement learning., IEEE COMPUTER SOC, 1998, INTERNATIONAL CONFERENCE ON MULTI-AGENT SYSTEMS, PROCEEDINGS, 479 - 480, web_of_science

    International conference proceedings

  • Interactive interfaces to detect conceptual difference for group knowledge acquisition

    S Nishida; T Yoshida; T Kondo

    Conceptual difference is a serious problem in group knowledge acquisition systems, especially when different people with different background participate in a group. This paper deals with conceptual difference and proposes a method to detect it in the cases that different symbols are used as the same meaning and/or same symbols are used as the different meanings. In section 2, conceptual difference is defined and system architecture for detection is described. In section 3, detecting algorithm is designed, and then a prototype system and its evaluation are discussed in section 4. Copyright (C) 1998 IFAC., PERGAMON PRESS LTD, 1998, ARTIFICIAL INTELLIGENCE IN REAL-TIME CONTROL 1997, 177 - 181, web_of_science

    International conference proceedings

  • 花紋スモッキングの展開図に基づく紐の生成

    吉田哲也; 藤田真奈美

    Mar. 2021, 芸術科学会論文誌, 20 (1), 10 - 20

MISC

  • Active Mining for Structured Data( Active Mining)

    Motoda Hiroshi; Ho Tu Bao; Washio Takashi; Yada Katsutoshi; Yoshida Tetsuya; Ohara Kouzou

    The Japanese Society for Artificial Intelligence, 01 Mar. 2005, Journal of Japanese Society for Artificial Intelligence, 20 (2), 172 - 179, cinii_articles

  • Extracting Diagnostic Knowledge from Hepatitis Data by Decision Tree Graph-Based Induction

    GEAMSAKUL Warodom; YOSHIDA Tetsuya; OHARA Kouzou; MOTODA Hiroshi; WASHIO Takashi

    Decision Tree Graph-Based Induction (DT-GBI) is a technique for constructing a decision tree from graph-structured data. In DT-GBI, substructures (discriminative patterns) are extracted by stepwise pair expansion (pair-wise chunking) and used as test attributes at nodes of a decision tree. We applied DT-GBI to a classification task of hepatitis data. In the first experiment, the stages of fibrosis are used as classes and a decision tree is constructed for discriminating patients with F4 (cirrhosis) from patients with the other stages using only the time sequence data of blood inspection. In..., Information Processing Society of Japan (IPSJ), 14 Sep. 2003, IPSJ SIG Notes. ICS, 2003 (90), 53 - 58, cinii_articles

  • Extracting Diagnostic Knowledge from Hepatitis Data by Decision Tree Graph-Based Induction (小特集 「アクティブマイニング」および一般)

    Geamsakul Warodom; 吉田 哲也; 大原 剛三

    人工知能学会, 14 Sep. 2003, 知識ベ-スシステム研究会, 61 (0), 53 - 58, cinii_articles

  • Extracting Diagnostic Knowledge from Hepatitis Data by Decision Tree Graph-Based Induction

    GEAMSAKUL Warodom; YOSHIDA Tetsuya; OHARA Kouzou; MOTODA Hiroshi; WASHIO Takashi

    Decision Tree Graph-Based Induction (DT-GBI) is a technique for constructing a decision tree from graph-structured data. In DT-GBI, substructures (discriminative patterns) are extracted by stepwise pair expansion (pair-wise chunking) and used as test attributes at nodes of a decision tree. We applied DT-GBI to a classification task of hepatitis data. In the first experiment, the stages of fibrosis are used as classes and a decision tree is constructed for discriminating patients with F4 (cirrhosis) from patients with the other stages using only the time sequence data of blood inspection. In..., The Institute of Electronics, Information and Communication Engineers, 07 Sep. 2003, IEICE technical report. Artificial intelligence and knowledge-based processing, 103 (304), 53 - 58, cinii_articles

  • MLnet

    吉田 哲也

    The Japanese Society for Artificial Intelligence, 01 Jul. 2003, Journal of Japanese Society for Artificial Intelligence, 18 (4), cinii_articles

  • Functional Extension of Decision Tree - Graph-Based Induction

    GEAMSAKUL Warodom; MATSUDA Takashi; YOSHIDA Tetsuya; MOTODA Hiroshi; WASHIO Takashi

    A machine learning technique called Graph-Based Induction (GBI) efficiently extracts typical patterns from graph-structured data by stepwise pair expansion (pairwaise chunking). Meanwhile, a decision tree is an effective means of data classification from which rules that are easy to understand can be obtained. However, a decision tree could not be produced for the data which is not explicitly expressed with attribute-value pairs. In this paper, we propose a method of constructing a classifier (decision tree) for graph-structured data by GBI. In our approach attributes, namely substructures ..., Information Processing Society of Japan (IPSJ), 13 Mar. 2003, IPSJ SIG Notes. ICS, 2003 (30), 93 - 98, cinii_articles

  • Correlation Analysis of 3-dimensional Chemical Structure and its Activity by AGM

    NISHIMURA Yoshio; WASHIO Takashi; YOSHIDA Tetsuya; MOTODA Hiroshi; INOKUCHI Akihiro; OKADA Takashi

    Apriori-based Graph Mining (AGM) algorithm efficiently extracts all the subgraph patterns which frequently appear in graph structured data. The algorithms can deal with general graph structured data with multiple labels of vartices and edges, and is capable of analyzing the connective structure of graphs. We have proposed a faster algorithm of AGM by adding an extra constraint to reduce the number of generated candidates for seeking frequent subgraphs. In this paper, we propose a new method to analyze graph structured data which are represented with a 3-dimensional coordinate by AGM. In thi..., Information Processing Society of Japan (IPSJ), 13 Mar. 2003, IPSJ SIG Notes. ICS, 2003 (30), 99 - 103, cinii_articles

  • Decision Tree-Graph-Based Inductionの機能拡張 (知識ベースシステム研究会(第60回) 人工知能基礎論研究会(第52回) 小特集:「データマイニング」および一般) -- (文部科学省科学研究費特定領域研究 情報洪水時代におけるアクティブマイニングの実現)

    Geamsakul Warodom; 松田 喬; 吉田 哲也

    人工知能学会, 13 Mar. 2003, 知識ベ-スシステム研究会, 60 (0), 93 - 98, cinii_articles

  • AGMによる3次元構造と生理活性の相関解析 (知識ベースシステム研究会(第60回) 人工知能基礎論研究会(第52回) 小特集:「データマイニング」および一般) -- (文部科学省科学研究費特定領域研究 情報洪水時代におけるアクティブマイニングの実現)

    西村 芳男; 鷲尾 隆; 吉田 哲也

    人工知能学会, 13 Mar. 2003, 知識ベ-スシステム研究会, 60 (0), 99 - 103, cinii_articles

  • Functional Extension of Decision Tree-Graph-Based Induction

    GEAMSAKUL Warodom; MATSUDA Takashi; YOSHIDA Tetsuya; MOTODA Hiroshi; WASHIO Takashi

    A machine learning technique called Graph-Based Induction (GBI) efficiently extracts typical patterns from graph-structured data by stepwise pair expansion (pairwaise chunking). Meanwhile, a decision tree is an effective means of data classification from which rules that are easy to understand can be obtained. However, a decision tree could not be produced for the data which is not explicitly expressed with attribute-value pairs. In this paper, we propose a method of constructing a classifier (decision tree) for graph-structured data by GBI. In our approach attributes, namely substructures ..., The Institute of Electronics, Information and Communication Engineers, 07 Mar. 2003, IEICE technical report. Artificial intelligence and knowledge-based processing, 102 (710), 41 - 46, cinii_articles

  • Correlation Analysis of 3-dimensional Chemical Structure and its Activity by AGM

    NISHIMURA Yoshio; WASHIO Takashi; YOSHIDA Tetsuya; MOTODA Hiroshi; INOKUCHI Akihiro; OKADA Takashi

    Apriori-based Graph Mining (AGM) algorithm efficiently extracts all the subgraph patterns which frequently appear in graph structured data. The algorithms can deal with general graph structured data with multiple labels of vartices and edges, and is capable of analyzing the connective structure of graphs. We have proposed a faster algorithm of AGM by adding an extra constraint to reduce the number of generated candidates for seeking frequent subgraphs. In this paper, we propose a new method to analyze graph structured data which are represented with a 3-dimensional coordinate by AGM. In thi..., The Institute of Electronics, Information and Communication Engineers, 07 Mar. 2003, IEICE technical report. Artificial intelligence and knowledge-based processing, 102 (710), 47 - 51, cinii_articles

  • グラフ構造データからのマイニング (特集 大阪大学産業科学研究所 マテリアルインテグレーション--材料・生体・情報の融合を目指して(2))

    元田 浩; 鷲尾 隆; 吉田 哲也

    ティ-・アイ・シ-, Aug. 2002, マテリアルインテグレ-ション, 15 (8), 59 - 62, cinii_articles

  • 専門家とデータの両方からの統一的知識獲得 (特集 大阪大学産業科学研究所 マテリアルインテグレーション--材料・生体・情報の融合を目指して(2))

    吉田 哲也; 元田 浩; 鷲尾 隆

    ティ-・アイ・シ-, Aug. 2002, マテリアルインテグレ-ション, 15 (8), 63 - 66, cinii_articles

  • A Faster Apriori-based Graph Algorithm

    Nishimura Yoshio; Washio Takashi; Yoshida Tetsuya; Motoda Hiroshi; Inokuchi Akihiro

    Apriori-based Graph Mining (AGM) algorithm derives suhgraph patterns efficiency which frequently appear in database consisting of graph structure data. In this paper, we propose a new and faster algorithm of ACM achieved by adding a condition to generate candidate frequent graphs. Simulation experiments were carried out to evaluate the performance of the proposed new ACM algorithm over the synthesized data and real world chemical data. Efficient reduction of the computation time by the proposed method has been confirmed., Information Processing Society of Japan (IPSJ), 23 May 2002, IPSJ SIG Notes. ICS, 2002 (45), 11 - 16, cinii_articles

  • Knowledge Discovery from Hepatitis Data based on Graph Structure

    Matsuda Takashi; Yoshida Tetsuya; Motoda Hiroshi; Washio Takashi

    Table representation is not suitable to represent data with many missing values in knowledge discovery, since the missing values are explicitly represented in the table and they can hinder the appropriate mining process. Graph structure can he robust for dealing with data with many missing values since they can just be omitted from the graph and thus are not explicitly represented. This paper reports a preliminary attempt to discover useful time sequence pattern from hepatitis data, which have many missing values, with Graph-Based Induction (GBI) method. Although some useful patterns for cl..., Information Processing Society of Japan (IPSJ), 23 May 2002, IPSJ SIG Notes. ICS, 2002 (45), 67 - 72, cinii_articles

  • Apriori-based Graph Mining アルゴリズムの高速化 (テーマ:「アクティブマイニング」および一般)

    西村 芳男; 鷲尾 隆; 吉田 哲也

    人工知能学会, 23 May 2002, 知識ベ-スシステム研究会, 56 (0), 11 - 16, cinii_articles

  • グラフ構造に着目した肝炎データからの知識発見 (テーマ:「アクティブマイニング」および一般)

    松田 喬; 吉田 哲也; 元田 浩

    人工知能学会, 23 May 2002, 知識ベ-スシステム研究会, 56 (0), 67 - 72, cinii_articles

  • Supporting Mutual Understanding in Participatoty Design Using Cases

    YAMAOKA Takayuki; TSUJINO Katsuhiko; YOSHIDA Tetsuya; NISHIDA Shogo

    ヒュ-マンインタフェ-ス学会, 01 May 2000, Human interface. The Transaction of Human Interface Society, 2 (2), 87 - 95, cinii_articles

  • Adaptive Hypermedia System for Supporting Information Providers to Direct Users through Hyperspace

    2000, 120-C (11), 1720 - 1731

  • Supporting Mutual Understanding in Participatory Design Using Cases

    2000, 2 (2), 87 - 95

  • Learning the balance between exploration and exploitation via reward

    T Yoshida; K Hori; S Nakasuka

    This paper proposes a new method to improve cooperation in concurrent systems within the framework of Multi-Agent Systems (MAS) by utilizing reinforcement learning. When subsystems work independently and concurrently, achieving appropriate cooperation among them is important to improve the effectiveness of the overall system. Treating subsystems as agents makes it easy to explicitly deal with the interactions among them since they can be modeled naturally as communication among agents with intended information. In our approach agents try to learn the appropriate balance between exploration and exploitation via reward, which is important in distributed and concurrent problem solving in general. By focusing on how to give reward in reinforcement learning, not the learning equation, two kinds of reward are defined in the context of cooperation between agents, in contrast to reinforcement learning within the framework of single agent. In our approach reward for insistence by individual agent contributes to facilitating exploration and reward for concessi:on to other agents contributes to facilitating exploitation. Our cooperation method was examined through experiments on the design of micro satellites and the result showed that it was effective to some extent to facilitate cooperation among agents by letting agents themselves learn the appropriate balance between insistence and concession. The result also suggested the possibility of utilizing the relative magnitude of these rewards as a new control parameter in MAS to control the overall behavior of MAS., IEICE-INST ELECTRONICS INFORMATION COMMUNICATIONS ENG, Nov. 1999, IEICE TRANSACTIONS ON FUNDAMENTALS OF ELECTRONICS COMMUNICATIONS AND COMPUTER SCIENCES, E82A (11), 2538 - 2545, web_of_science

  • 決定木間のリンクを利用した概念相違発見手法 (テーマ:「インターネットとAI」および一般)

    大西 健介; 吉田 哲也; 西田 正吾

    人工知能学会, Mar. 1999, 知識ベ-スシステム研究会, 69 - 74, cinii_articles

  • 検索キーワードの補完情報を利用した情報獲得支援 (テーマ:「インターネットとAI」および一般)

    新開 大樹; 吉田 哲也; 西田 正吾

    人工知能学会, Mar. 1999, 知識ベ-スシステム研究会, 105 - 110, cinii_articles

  • Discovering Conceptual Defferences among Different People via Diverse Structures

    1999, The Third Pacific-Asia Conference on Knowledye Discovery and Data Mining(PAKDD-99), 494 - 498

  • A Communication Model in Emergency which Considers Competence, Duty, Responsiblity and Knowledge

    1999, 82A (3), 445 - 453

  • A Dynamic Linkage Method for Hypermedia Based on Metadata and User Model

    Hijikata Yoshinori

    Information Processing Society of Japan (IPSJ), 05 Oct. 1998, 全国大会講演論文集, 57 (3), 105 - 106, cinii_articles

  • A cooperation method via metaphor of explanation

    T Yoshida; K Hori; S Nakasuka

    This paper proposes a new method to improve cooperation in concurrent systems within the framework of Multi-Agent Systems (MAS). Since subsystems work concurrently, achieving appropriate cooperation among them is important to improve the effectiveness of the overall system. When subsystems are modeled as agents, it is easy to explicitly deal with the interactions among them since they can be modeled naturally as communication among agents with intended information. Contrary to previous approaches which provided the syntax of communication protocols without semantics, we focus on the semantics of cooperation in MAS and aim at allowing agents to exploit the communicated information for cooperation. This is attempted by utilizing more coarse-grained communication based on the different perspective for the balance between Formality and richness of communication contents so that each piece of communication contents can convey more meaningful information in application domains. In our approach agents cooperate each other by giving Feedbacks based on the metaphor of explanation which is widely used in human interactions, in contrast to previous approaches which use direct orders given by the leader based on the pre-defined cooperation strategies. Agents show the difference between the proposal and counterproposals for it, which are constructed with respect to the Former and given as the feedbacks in the easily understandable terms For the receiver. From the comparison of proposals agents retrieve the information on which parts are agreed and disagreed by the relevant agents, and reflect the analysis in their following behavior. Furthermore, communication contents are annotated by agents to indicate the degree of importance in decision making for them, which contributes to making explanations or feedbacks more understandable. Our cooperation method was examined through experiments on the design of micro satellites and the result showed that it was effective to some extent to facilitate cooperation among agents., IEICE-INST ELECTRONICS INFORMATION COMMUNICATIONS ENG, Apr. 1998, IEICE TRANSACTIONS ON FUNDAMENTALS OF ELECTRONICS COMMUNICATIONS AND COMPUTER SCIENCES, E81A (4), 576 - 585, web_of_science

  • Software Design Support by Utilizing Comment

    1998, 113 - 118

  • Facilitate Cooperation of Humans and Machines from the viewpoint of Multi-Agent Systems

    1998, 7th IFAC SYMPOSIUM on Analysis, Design and Evaluation of Man-Machine Systems, 485 - 490

  • Discovering conceptual differences among people from cases

    T Yoshida; T Kondo

    We propose a method for discovering conceptual differences (CD) among people from cases. In general different people seem to have different ways of conception and thus can have different concepts even on the same thing. Removing CD seems especially important when people with different backgrounds and knowledge carry out collaborative works as a group; otherwise they cannot communicate ideas and establish mutual understanding even on the same thing. In our approach knowledge of users is structured into decision trees so that differences in concepts can be discovered as the differences in the structure of trees. Based on the candidates suggested by the system with our discovering algorithms, the users then discuss each other on differences in their concepts and modify them to reduce the differences. CD is gradually removed by repeating the interaction between the system and users. Experiments were carried out on the cases for motor diagnosis with artificially encoded CD. Admittedly our approach is simple, however, the result shows that our approach is effective to some extent as the first step toward dealing with the issue of CD among people., SPRINGER-VERLAG BERLIN, 1998, DISCOVERY SCIENCE, 1532, 162 - 173, web_of_science

  • An Extension of Reinforcement Learning in Multi-Agent Systems

    1998, 7th IEEE International Workshop on Robot and Human Communication(ROMAN'98), 221 - 226

  • A cooperation method to exploit design rationales for agents

    T Yoshida; K Hori; S Nakasuka

    This paper proposes a new method to improve cooperation in Multi-Agent Systems (MAS) from the viewpoint of Design Rationale (DR). In our approach agents cooperate each other by giving feedbacks on the proposal from each agent indirectly, in contrast to previous approaches which use direct orders given by the leader based on the pre-defined cooperation strategies. Indirect cooperation is realized by constructing and giving counter-proposals from relevant agents on the proposal. Feedbacks given as the counter-proposals play the role of pointing out insufficient portions in the original one from different perspectives toward the agreement among agents. Based on the comparison of proposals agents retrieve the information on which parts are agreed and disagreed by the relevant agents, and reflect the analysis in their following behavior. Furthermore, communication contents are annotated by agents to indicate the degree of importance in decision making for them, which contributes to making feedbacks more understandable. Our cooperation method was examined through experiments on the design of micro satellites and the result showed that it was effective to some extent to facilitate cooperation among agents., I O S PRESS, 1998, INTELLIGENT AUTONOMOUS SYSTEMS, 647 - 654, web_of_science

  • Supporting mutual understanding in collaborative design project

    T Yamaoka; K Tsujino; T Yoshida; S Nishida

    In collaborative design, participants usually have different backgrounds and standpoints. Due to the differences, it is hard for the participants mutually to understand the intention and they have to make much effort to reach a mutual agreement on the design, then the efficiency of design processes is decreased. In this paper, we propose a framework to support the participants for making the mutual understanding by grasping and showing differences among their intention in collaboration. A system based on the proposed framework has the characteristics: facilitating mutual understanding among participants by presuming their intention and by communicating them with the understandable form, visualizing the difference in concepts among participants, and making progress collaboration in an interactive way., IEEE COMPUTER SOC, 1998, 3RD ASIA PACIFIC COMPUTER HUMAN INTERACTION, PROCEEDINGS, 132 - 137, web_of_science

  • Design of the Interfaces to Detect Conceptual Differnce among Defferent People

    1998, 39 (5), 1195 - 1202

  • Evaluation of organizational structure in emergency from the viewpoint of communication

    Koiso Takashi; Saiwaki Naoki; Yoshida Tetsuya; Nishida Shogo

    It is very important to support communications in the emergency situation for large scale systems. We focus on the evaluation of the organizational structure for emergent situations from the viewpoint of communication. Our approach emphasize quantitative analysis which contracts to the qualitative analysis in social science. First, we explain communication model briefly, and discuss how to evaluate organizations by using the model. Furthermore some prototype system are explained., The Institute of Electronics, Information and Communication Engineers, 13 Jun. 1997, Technical report of IEICE. HCS, 97 (99), 5 - 12, cinii_articles

  • Coloured rippling : an extension of a theorem proving heuristics

    1994, 11th European Conference on Artificial Intelligence, 85 - 89

Presentations

  • リジッドルームのための織物組織図の近似学習

    吉田哲也

    情報処理学会 第132回数理モデル化と問題解決研究発表会, 01 Mar. 2021, 01 Mar. 2021, 02 Mar. 2021

  • 正多角形の貼り合わせを用いた花紋スモッキングの組み合わせの拡張

    夛田美沙; 吉田哲也

    NICOGRAPH2020, 03 Nov. 2020, 01 Nov. 2020, 03 Nov. 2020

Association Memberships

  • 日本建築学会

  • 芸術科学会

  • 情報処理学会

  • 人工知能学会



Copyright © MEDIA FUSION Co.,Ltd. All rights reserved.