search for




 

Discovering a fine dust pathway via directed acyclic graphical models
Journal of the Korean Data & Information Science Society 2019;30:67-76
Published online January 31, 2019;  https://doi.org/10.7465/jkdi.2019.30.1.67
© 2019 Korean Data and Information Science Society.

Gunwoong Park1

1Department of Statistics, University of Seoul
Correspondence to: Assistant professor, Department of Statistics, University of Seoul, Seoul 02504, Korea. E-mail: gw.park23@gmail.com
Received December 27, 2018; Revised January 15, 2019; Accepted January 17, 2019.
This is an Open Access article distributed under the terms of the Creative Commons Attribution Non-Commercial License (http://creativecommons.org/licenses/by-nc/3.0) which permits unrestricted non-commercial use, distribution, and reproduction in any medium, provided the original work is properly cited.
Abstract
The main objective of this paper is to analyze the fine dust (PM-10) directed pathway in Seoul using directed acyclic graphical (DAG) models. In this paper, a Gaussian DAG model is applied which is one of the most widely used for recovering the underlying structure of multivariate continuous data. Among a number of constraint-based and score-based algorithms for learning directed graphs, we exploit the hybrid max-min hill climbing (MMHC) algorithm where both constraint-based and score-based approaches are applied. We verify through the 2017 Seoul fine dust data that our method is well suited for estimating the partial fine dust pathway where it is consistent to seasonal wind directions. We expect that the estimated fine dust pathway can be exploited for various fine dust reduction methodologies.
Keywords : Bayesian network, directed acyclic graphical model, fine dust pathway, monsoon.