Logistic Regression in the Identification of Hazards in Construction

World Multidisciplinary Civil Engineering-Architecture-Urban Planning Symposium - WMCAUS

土木建筑工程

Drozd, Wojciech^1

Institute of Building and Transport Management, Tadeusz Kociuszko Cracow University of Technology, Warszawska 24 St., Kraków

31-155, Poland^1

关键词: Accidents at work;    Causes of accidents;    Construction sites;    Logistic regressions;    Qualitative characteristics;    Statistical classifier;   

Others  :  https://iopscience.iop.org/article/10.1088/1757-899X/245/6/062012/pdf
DOI  :  10.1088/1757-899X/245/6/062012

学科分类：土木及结构工程学

来源: IOP

【 摘 要 】

The construction site and its elements create circumstances that are conducive to the formation of risks to safety during the execution of works. Analysis indicates the critical importance of these factors in the set of characteristics that describe the causes of accidents in the construction industry. This article attempts to analyse the characteristics related to the construction site, in order to indicate their importance in defining the circumstances of accidents at work. The study includes sites inspected in 2014 - 2016 by the employees of the District Labour Inspectorate in Krakow (Poland). The analysed set of detailed (disaggregated) data includes both quantitative and qualitative characteristics. The substantive task focused on classification modelling in the identification of hazards in construction and identifying those of the analysed characteristics that are important in an accident. In terms of methodology, resource data analysis using statistical classifiers, in the form of logistic regression, was the method used.