Mining Multi-scale Intervention Rules from Time Series and Complex Network

DOI

10.2991/ijcis.2011.4.4.25How to use a DOI?

Keywords

Intervention rule, Complex network, Time series, Decomposition, Multi-scale

Abstract

This paper proposes the concept of intervention rule which tries to reveal the interventional relationship between elements in a system in the following three aspects. (1) Casual relationship. Intervention rule shows which element is the cause and which element is the consequence. (2) Quantitative relationship: Intervention rule shows the quantitative intensity of how the change of the causal element interferes with the change of the consequential element. (3) Multi-scale intervention relationship. Intervention rule shows the intervention at different decomposition scale of the original system, since sub system may exhibit different mechanism from the original system. This paper first introduces a general intervention rule framework, and then transforms the framework into concrete intervention rules for complex network data and time series data. Then, it proposes two algorithms to mine the intervention rules from the two different systems. Finally, the experimental results show that multi-scale intervention rules do exist in real dataset. And the intervention intensity of each sub graph and sub series are always 4 or 5 time larger than intervention intensity of the original data.

Copyright

© 2011, the Authors. Published by Atlantis Press.

Open Access

This is an open access article distributed under the CC BY-NC license (http://creativecommons.org/licenses/by-nc/4.0/).

Download article (PDF)

TY  - JOUR
AU  - Jiaoling Zheng
AU  - Changjie Tang
AU  - Shaojie Qiao
AU  - Ning Yang
AU  - Yue Wang
PY  - 2011
DA  - 2011/09/02
TI  - Mining Multi-scale Intervention Rules from Time Series and Complex Network
JO  - International Journal of Computational Intelligence Systems
SP  - 728
EP  - 738
VL  - 4
IS  - 4
SN  - 1875-6883
UR  - https://doi.org/10.2991/ijcis.2011.4.4.25
DO  - 10.2991/ijcis.2011.4.4.25
ID  - Zheng2011
ER  -