@ARTICLE{Krawczyńska-Piechna_Anna_Predicting,
 author={Krawczyńska-Piechna, Anna},
 volume={Vol. 66},
 number={No 3},
 journal={Archives of Civil Engineering},
 pages={365-376},
 howpublished={online},
 publisher={WARSAW UNIVERSITY OF TECHNOLOGY FACULTY OF CIVIL ENGINEERING and COMMITTEE FOR CIVIL ENGINEERING POLISH ACADEMY OF SCIENCES},
 abstract={Work safety control and analysis of accidents during the construction performance are some of the most important issues of the construction management. The paper focuses on the post-accident absence as an element of the occupational safety management. The occurrence of the post-accident absence of workers can be then treated as an indicator of building performance safety. The ability to estimate its length can also facilitate works planning and scheduling in case of the accident. The paper attempts to answer the question whether it is possible and how to use decision trees and their ensembles to predict the severity of the post-accident absence and which classification algorithm is the most promising to solve the prediction problem. The paper clarifies the model of the prediction problem, introduces 5 different decision tress and different aggregation algorithms in order to build the model. Thanks to the use of aggregation methods it is possible to build classifiers that predict precisely and do not require any initial data treatment, which simplifies the prediction process significantly. To identify the most promising classifier or classifier ensemble the prediction accuracy measures of selected classification algorithms were analyzed. The data to build the model was gathered on national (Polish) construction sites and was taken from literature. Models obtained within simulations can be used to build advisory or safety management systems allowing to detect threats while construction works are being planned or carried out.},
 type={Article},
 title={Predicting the Length of a Post-Accident Absence in Construction with Decision Trees and Their Ensembles},
 URL={http://journals.pan.pl/Content/117469/PDF/20.ACE00013%20do%20druku_B5.pdf},
 doi={10.24425/ace.2020.134402},
 keywords={post-accident absence, decision trees, prediction, classification ensembles},
}