Journal of Japan Society of Civil Engineers, Ser. D3 (Infrastructure Planning and Management)
Online ISSN : 2185-6540
ISSN-L : 2185-6540
Infrastructure Planning and Management Vol.31 (Special Issue)
VISUALIZATION OF INTER-REGIONAL DEPENDENCE BASED ON OD DATA BY USING BAYESIAN NETWORK
Takashi FUSETakuya WATANABE
Author information
JOURNAL FREE ACCESS

2014 Volume 70 Issue 5 Pages I_55-I_61

Details
Abstract

Inter-regional dependence of traveling is important to understand the regional characteristics. So far, various statistics have been prepared for grasping the regions. Recently, extraction methods of dependence between variables using Bayesian network have been developed. This study proposes visualization method of inter-regional dependence based on OD data by using the Bayesian network. The Bayesian network expresses the dependence between variables as directed graph with conditional probability. The conditional probability is modeled by B-spline non-linear regression model. The graph structure is inferred based on maximum a posteriori probability. We apply the proposed method to the Inter-Regional Travel Survey in Japan, and attempt to visualize inter-regional dependence.

Content from these authors
© 2014 Japan Society of Civil Engineers
Previous article Next article
feedback
Top