NORMA eResearch @NCI Library

Predicting Fatalities: Enhancing Construction Site Safety Through Advanced Machine Learning

Kirola, Deepak Singh (2024) Predicting Fatalities: Enhancing Construction Site Safety Through Advanced Machine Learning. Masters thesis, Dublin, National College of Ireland.

[thumbnail of Master of Science]
Preview
PDF (Master of Science)
Download (1MB) | Preview
[thumbnail of Configuration Manual]
Preview
PDF (Configuration Manual)
Download (2MB) | Preview

Abstract

In the construction industry, ensuring workplace safety is a critical challenge. Jaby Mohammed and Md Jubaer Mahmud, from the Department of Technology at Illinois State University, address this issue by analyzing Occupational Safety and Health Administration (OSHA) Accident and Injury Data from 2015 to 2017. They use machine learning techniques, exploring nine algorithms, including XGBoost and Random Forest, which achieve accuracies of 65.29% and 58.24%, with AUC values of 78.83% and 69.52%, respectively. To improve their model, this research pays attention to specific details in the data and fine-tunes their methods, focusing on precision. After this adjustment, this research achieves an accuracy of 85.7% and an AUC of 92.5% for XGBoost, and an accuracy of 84.9% and an AUC of 91.3% for AdaBoost. This signifies that the methods employed in this research are more effective in predicting outcomes.

The study aims to transform safety management in the construction industry by establishing a data-driven system. By uncovering injury patterns, causal factors, and areas requiring improved safety measures in the unstructured OSHA data, the research contributes significantly to the workplace safety literature. This work envisions a safer future by combining advanced machine learning methods, detailed data analysis, and refined predictive models. The insights from this research offer practical guidance for safety-conscious organizations, with the potential to positively impact workplace safety practices and decision-making.

Item Type: Thesis (Masters)
Supervisors:
Name
Email
Moldovan, Arghir-Nicolae
UNSPECIFIED
Subjects: Q Science > QA Mathematics > Electronic computers. Computer science
T Technology > T Technology (General) > Information Technology > Electronic computers. Computer science
H Social Sciences > HD Industries. Land use. Labor > Specific Industries > Construction Industry
H Social Sciences > HD Industries. Land use. Labor > Issues of Labour and Work > Health and Safety at Work.
Q Science > Q Science (General) > Self-organizing systems. Conscious automata > Machine learning
Divisions: School of Computing > Master of Science in Data Analytics
Depositing User: Ciara O'Brien
Date Deposited: 05 Jun 2025 11:57
Last Modified: 05 Jun 2025 11:57
URI: https://norma.ncirl.ie/id/eprint/7756

Actions (login required)

View Item View Item