Intrinsic graph structure estimation using graph Laplacian

Atsushi Noda, Hideitsu Hino, Masami Tatsuno, Shotaro Akaho, Noboru Murata

    Research output: Contribution to journalArticle

    4 Citations (Scopus)

    Abstract

    A graph is a mathematical representation of a set of variables where some pairs of the variables are connected by edges. Common examples of graphs are railroads, the Internet, and neural networks. It is both theoretically and practically important to estimate the intensity of direct connections between variables. In this study, a problem of estimating the intrinsic graph structure from observed data is considered. The observed data in this study are a matrix with elements representing dependency between nodes in the graph. The dependency representsmore than direct connections because it includes influences of various paths. For example, each element of the observed matrix represents a co-occurrence of events at two nodes or a correlation of variables corresponding to two nodes. In this setting, spurious correlations make the estimation of direct connection difficult. To alleviate this difficulty, a digraph Laplacian is used for characterizing a graph. A generative model of this observed matrix is proposed, and a parameter estimation algorithm for the model is also introduced.The notable advantage of the proposedmethod is its ability to deal with directed graphs,while conventional graph structure estimation methods such as covariance selections are applicable only to undirected graphs. The algorithm is experimentally shown to be able to identify the intrinsic graph structure.

    Original languageEnglish
    Pages (from-to)1455-1483
    Number of pages29
    JournalNeural Computation
    Volume26
    Issue number7
    DOIs
    Publication statusPublished - 2014

    Fingerprint

    Railroads
    Internet
    Graph
    Intrinsic

    ASJC Scopus subject areas

    • Cognitive Neuroscience
    • Arts and Humanities (miscellaneous)

    Cite this

    Intrinsic graph structure estimation using graph Laplacian. / Noda, Atsushi; Hino, Hideitsu; Tatsuno, Masami; Akaho, Shotaro; Murata, Noboru.

    In: Neural Computation, Vol. 26, No. 7, 2014, p. 1455-1483.

    Research output: Contribution to journalArticle

    Noda, Atsushi ; Hino, Hideitsu ; Tatsuno, Masami ; Akaho, Shotaro ; Murata, Noboru. / Intrinsic graph structure estimation using graph Laplacian. In: Neural Computation. 2014 ; Vol. 26, No. 7. pp. 1455-1483.
    @article{8547da981b8f46d69964740b7530ff97,
    title = "Intrinsic graph structure estimation using graph Laplacian",
    abstract = "A graph is a mathematical representation of a set of variables where some pairs of the variables are connected by edges. Common examples of graphs are railroads, the Internet, and neural networks. It is both theoretically and practically important to estimate the intensity of direct connections between variables. In this study, a problem of estimating the intrinsic graph structure from observed data is considered. The observed data in this study are a matrix with elements representing dependency between nodes in the graph. The dependency representsmore than direct connections because it includes influences of various paths. For example, each element of the observed matrix represents a co-occurrence of events at two nodes or a correlation of variables corresponding to two nodes. In this setting, spurious correlations make the estimation of direct connection difficult. To alleviate this difficulty, a digraph Laplacian is used for characterizing a graph. A generative model of this observed matrix is proposed, and a parameter estimation algorithm for the model is also introduced.The notable advantage of the proposedmethod is its ability to deal with directed graphs,while conventional graph structure estimation methods such as covariance selections are applicable only to undirected graphs. The algorithm is experimentally shown to be able to identify the intrinsic graph structure.",
    author = "Atsushi Noda and Hideitsu Hino and Masami Tatsuno and Shotaro Akaho and Noboru Murata",
    year = "2014",
    doi = "10.1162/NECO_a_00603",
    language = "English",
    volume = "26",
    pages = "1455--1483",
    journal = "Neural Computation",
    issn = "0899-7667",
    publisher = "MIT Press Journals",
    number = "7",

    }

    TY - JOUR

    T1 - Intrinsic graph structure estimation using graph Laplacian

    AU - Noda, Atsushi

    AU - Hino, Hideitsu

    AU - Tatsuno, Masami

    AU - Akaho, Shotaro

    AU - Murata, Noboru

    PY - 2014

    Y1 - 2014

    N2 - A graph is a mathematical representation of a set of variables where some pairs of the variables are connected by edges. Common examples of graphs are railroads, the Internet, and neural networks. It is both theoretically and practically important to estimate the intensity of direct connections between variables. In this study, a problem of estimating the intrinsic graph structure from observed data is considered. The observed data in this study are a matrix with elements representing dependency between nodes in the graph. The dependency representsmore than direct connections because it includes influences of various paths. For example, each element of the observed matrix represents a co-occurrence of events at two nodes or a correlation of variables corresponding to two nodes. In this setting, spurious correlations make the estimation of direct connection difficult. To alleviate this difficulty, a digraph Laplacian is used for characterizing a graph. A generative model of this observed matrix is proposed, and a parameter estimation algorithm for the model is also introduced.The notable advantage of the proposedmethod is its ability to deal with directed graphs,while conventional graph structure estimation methods such as covariance selections are applicable only to undirected graphs. The algorithm is experimentally shown to be able to identify the intrinsic graph structure.

    AB - A graph is a mathematical representation of a set of variables where some pairs of the variables are connected by edges. Common examples of graphs are railroads, the Internet, and neural networks. It is both theoretically and practically important to estimate the intensity of direct connections between variables. In this study, a problem of estimating the intrinsic graph structure from observed data is considered. The observed data in this study are a matrix with elements representing dependency between nodes in the graph. The dependency representsmore than direct connections because it includes influences of various paths. For example, each element of the observed matrix represents a co-occurrence of events at two nodes or a correlation of variables corresponding to two nodes. In this setting, spurious correlations make the estimation of direct connection difficult. To alleviate this difficulty, a digraph Laplacian is used for characterizing a graph. A generative model of this observed matrix is proposed, and a parameter estimation algorithm for the model is also introduced.The notable advantage of the proposedmethod is its ability to deal with directed graphs,while conventional graph structure estimation methods such as covariance selections are applicable only to undirected graphs. The algorithm is experimentally shown to be able to identify the intrinsic graph structure.

    UR - http://www.scopus.com/inward/record.url?scp=84902207976&partnerID=8YFLogxK

    UR - http://www.scopus.com/inward/citedby.url?scp=84902207976&partnerID=8YFLogxK

    U2 - 10.1162/NECO_a_00603

    DO - 10.1162/NECO_a_00603

    M3 - Article

    C2 - 24708372

    AN - SCOPUS:84902207976

    VL - 26

    SP - 1455

    EP - 1483

    JO - Neural Computation

    JF - Neural Computation

    SN - 0899-7667

    IS - 7

    ER -