工藤 峰一 Kudo M.

ID:1000060205101

北海道大学大学院情報科学研究科 Graduate School of Information Science Technology, Hokkaido University (2013年 CiNii収録論文より)

Search authors sharing the same name

Articles:  21-40 of 74

  • 2 / 4
  • Complexity and Enumeration of Subclasses  [in Japanese]

    NAKAMURA Atsuyoshi , KUDO Mineichi

    Representing a complicated region by union of simple regions is often preferred in various study areas because each component simple region is easy to understand and easy to deal with. Kudo et al. [2] …

    IEICE technical report 107(219), 35-42, 2007-09-13

    References (6)

  • A Study on Sitting-Posture Analysis by Pressure Sensors  [in Japanese]

    KAMIYA Kazuhiro , KUDO Mineichi , NONAKA Hidetoshi , TOYAMA Jun

    It has been promising to provide customized services for improving quality in support of information technologies. Sitting is one of natural actions in our life. We focus on the sittig behavior as a c …

    情報処理学会研究報告. UBI, [ユビキタスコンピューティングシステム] 15, 41-46, 2007-07-19

    References (11)

  • A Study on Sitting-Posture Analysis by Pressure Sensors  [in Japanese]

    KAMIYA Kazuhiro , KUDO Mineichi , NONAKA Hidetoshi , TOYAMA Jun

    It has been promising to provide customized services for improving quality in support of information technologies. Sitting is one of natural actions in our life. We focus on the sittig behavior as a c …

    IEICE technical report 107(152), 41-46, 2007-07-12

    References (11)

  • Analysis of Volume Prototypes and Comparison with Mixture Models  [in Japanese]

    SATO Maiko , KUDO Mineichi , TOYAMA Jun

    In these years, we often deal with an enormous amount of data in a large variety of pattern recognition tasks. As an efficient prototype, we have proposed a volume prototype. In this paper, we analyze …

    IEICE technical report 107(114), 93-98, 2007-06-28

    References (4)

  • Classifier Fusion on the Basis of Data Selection and Feature Selection  [in Japanese]

    SHIRAI Satoshi , KUDO Mineichi

    In recent years, many approaches to gain high performance by combining some classifiers have been proposed. In the bagging, we exploit many random replicates of samples, and in the random subspace met …

    IEICE technical report 107(114), 69-74, 2007-06-28

    References (10)

  • Finding of Frequent Similar Subtrees in Tree-Structured Data  [in Japanese]

    TOSAKA Hisashi , NAKAMURA Atsuyoshi , KUDO Mineichi

    We study a novel problem of mining subtrees with frequent occurrence of similar subtrees, and propose an efficient algorithm for this problem. According to our problem setting, frequency of a subtree …

    IEICE technical report 107(114), 7-12, 2007-06-28

    References (9)

  • Analysis of Volume Prototypes and Comparison with Mixture Models  [in Japanese]

    SATO Maiko , KUDO Mineichi , TOYAMA Jun

    In these years, we often deal with an enormous amount of data in a large variety of pattern recognition tasks. As an efficient prototype, we have proposed a volume prototype. In this paper, we analyze …

    IEICE technical report 107(115), 93-98, 2007-06-21

    References (4)

  • Classifier Fusion on the Basis of Data Selection and Feature Selection  [in Japanese]

    SHIRAI Satoshi , KUDO Mineichi

    In recent years, many approaches to gain high performance by combining some classifiers have been proposed. In the bagging, we exploit many random replicates of samples, and in the random subspace met …

    IEICE technical report 107(115), 69-74, 2007-06-21

    References (10)

  • Finding of Frequent Similar Subtrees in Tree-Structured Data  [in Japanese]

    TOSAKA Hisashi , NAKAMURA Atsuyoshi , KUDO Mineichi

    We study a novel problem of mining subtrees with frequent occurrence of similar subtrees, and propose an efficient algorithm for this problem. According to our problem setting, frequency of a subtree …

    IEICE technical report 107(115), 7-12, 2007-06-21

    References (9)

  • Information Compression by Volume Prototypes  [in Japanese]

    TABATA Kenji , KUDO Mineichi

    In these years, we often deal with an enormous amount of data in a large variety of pattern recognition tasks. Such data require a huge amount of memory space and computation time for processing. One …

    IEICE technical report 106(429), 25-30, 2006-12-08

    References (21) Cited by (2)

  • An Attribute Value Abstraction Using Granularity  [in Japanese]

    MUTO Yuji , KUDO Mineichi

    Generally, real data are given with several formats in the fields of pattern recognition and data mining. Therefore, discretization of continuous attributes and grouping of categorical attributes are …

    IEICE technical report 106(429), 19-24, 2006-12-08

    References (13)

  • Feature Selection : So far and from now on  [in Japanese]

    KUDO Mineichi

    Feature Selection is one of the most important techniques in pattern recognition. We will survey several methods presented so far and introduce new trials under the increasing demands of scalability a …

    IEICE technical report 106(428), 37-42, 2006-12-07

    References (19)

  • 繰返し構造をもつラベル付順序木の簡潔な表現法 (計算理論とアルゴリズムの新展開 RIMS研究集会報告集)  [in Japanese]

    Saito Tomoya , Nakamura Atsuyoshi , Kudo Mineichi

    RIMS Kokyuroku (1489), 216-222, 2006-05

    IR 

  • Detection of Wrong Characters by Probability Transitional Patterns of Two-Directional N-gram Probabilities  [in Japanese]

    KAWATA Takehiro , KUDO Mineichi , TOYAMA Jun , NAKAMURA Atsuyoshi

    OCRなどを通して得られる日本語文の認識結果において, N-gram確率を利用した高速な誤認識文字検出法を提案する.日本語のように単語が分かち書きされず大規模な語彙を対象とした場合, 誤り個所の指摘に文字N-gramは有効な方法である.本論文ではまず, 通常のN-gram確率の拡張として両方向N-gram確率を提案し, その有効性を情報量の点から考察する.次に, 両方向N-gram確率と文脈確率を …

    The IEICE transactions on information and systems Pt. 2 88(3), 629-635, 2005-03-01

    References (14)

  • Nonparametric Classification by Covering Using Family of Minimum Enclosing Spheres  [in Japanese]

    TAKIGAWA Ichigaku , KUDO Mineichi

    We propose a nonparametric multi-class classifier based on a family of spheres, each of which is the minimum covering sphere for a subset of positive samples and does not contain any negative samples. …

    Technical report of IEICE. PRMU 104(524), 37-42, 2004-12-17

    References (21)

  • Mining Frequent Trees with Node-Inclusion Constraints

    NAKAMURA Atsuyoshi , KUDO Mineichi

    In this paper, we propose an efficient algorithm enumerating all frequent subtrees containing all special nodes that are guaranteed to be included in all trees belonging to a given data. Our algorithm …

    IEICE technical report. Theoretical foundations of Computing 104(319), 7-14, 2004-10-15

    References (9)

  • Mining Frequent Trees with Node-Inclusion Constraints

    NAKAMURA Atsuyoshi , KUDO Mineichi

    In this paper, we propose an efficient algorithm enumerating all frequent subtrees containing all special nodes that are guaranteed to be included in all trees belonging to a given data. Our algorithm …

    IPSJ SIG Notes 97, 67-74, 2004-10-14

    References (9)

  • Effective Construction of Minimum -Norm Sequences on Underdetermined Signal Recovery  [in Japanese]

    TAKIGAWA Ichigaku , TOYAMA Jun , KUDO Mineichi

    On the problem like underdetermind signal separation of speech signal mixtures, that is, separation of more sources than mixtures, we need to find a unique and practically useful solution after the li …

    Technical report of IEICE. PRMU 103(296), 113-118, 2003-09-09

    References (16)

  • Acceleration of the κ-nearest neighbor method using inclusion and exclusion  [in Japanese]

    KUDO Mineichi

    The termination conditions were previously derived by us to stop the search procedure for finding the κ-nearest neighbors and a branch and bound algorithm equipped with these conditions was proposed. …

    Technical report of IEICE. PRMU 103(295), 91-95, 2003-09-08

    References (10) Cited by (1)

  • Relationship between Rule Accuracy and a Degree of Interest  [in Japanese]

    SHIDARA Yohji , NAKAMURA Atsuyoshi , KUDO Mineichi

    The methods which extract rules from database are useful for mining because of their human-readable output. Various methods have been proposed which construct some accurate classifiers and extract the …

    Technical report of IEICE. PRMU 103(295), 6-11, 2003-09-08

    References (4)

  • 2 / 4
Page Top