Citation

Lu S, Zhang D, Liu C. Procedia Eng. 2018; 211: 521-530.

Copyright

(Copyright © 2018, Elsevier Publishing)

DOI

10.1016/j.proeng.2017.12.044

PMID

unavailable

Abstract

To obtain the provincial fire characteristics in China from 2002 to 2010, a principal component analysis (PCA) was used to compress and interpret the variables. Eight fire-related variables were divided into two data sets. The general fire situation set consisted of the number of fires, fire deaths, direct loss and number of high-casualty fires (HCFs), and the fatal fire data set included the number of fires with fatalities of 1-2, 3-5, 6-9 and ≥10. The provincial fire characteristics were measured by the principal components that could reflect the fire situation, the ratio of general fires to HCFs, the fatal fire situation and the ratio of high fatality fires to low fatality fires. The provinces with similar PCA results were clustered together by clustering analysis. Guangdong, Zhejiang, Jiangsu and Henan are likely to experience serious fire situations and disastrous fires. The associations between the PCA results and the five influencing factors, such as GDP, per capita GDP, population, fire stations per million people and time, were analyzed by means of the relative risk (RR).

Language: en

Keywords

clustering analysis; fire statistics; influencing factors; principal component analysis; provincial characteristics; relative risk