2014 | OriginalPaper | Buchkapitel
An Algorithm for Directed Graph Estimation
verfasst von : Hideitsu Hino, Atsushi Noda, Masami Tatsuno, Shotaro Akaho, Noboru Murata
Erschienen in: Artificial Neural Networks and Machine Learning – ICANN 2014
Verlag: Springer International Publishing
Aktivieren Sie unsere intelligente Suche, um passende Fachinhalte oder Patente zu finden.
Wählen Sie Textabschnitte aus um mit Künstlicher Intelligenz passenden Patente zu finden. powered by
Markieren Sie Textabschnitte, um KI-gestützt weitere passende Inhalte zu finden. powered by
A problem of estimating the intrinsic graph structure from observed data is considered. The observed data in this study is a matrix with elements representing
dependency
between nodes in the graph. Each element of the observed matrix represents, for example, co-occurrence of events at two nodes, or correlation of variables corresponding to two nodes. The dependency does not represent direct connections and includes influences of various paths, and spurious correlations make the estimation of direct connection difficult. To alleviate this difficulty, digraph Laplacian is used for characterizing a graph. A generative model of an observed matrix is proposed, and a parameter estimation algorithm for the model is also proposed. The proposed method is capable of dealing with directed graphs, while conventional graph structure estimation methods from an observed matrix are only applicable to undirected graphs. Experimental result shows that the proposed algorithm is able to identify the intrinsic graph structure.