Abstract
Project contingency is one of the most important preliminary processes in any project estimation as it ensures the project risk for cost over-run due to uncertainties , in light of that, it is also important to make sure that the project contingency is not overestimated to maintain competitiveness in the market. This project aims to develop contingency decision support tool using ML techniques in order to predict the optimum contingency cost that balances between maintaining business competitiveness in the market and achieving project objectives. Different ML ways were evaluated and based on accuracy level ; Random Forest was found to be yielding the most accurate results using knime software. The required data to build the model are collected from organization’s database including a total of 2071 projects. In addition to interviews and surveys with project managers, the model covered the risk attribute, project value, line of business to optimize the project contingency cost.
Publication Date
12-10-2023
Document Type
Master's Project
Student Type
Graduate
Degree Name
Professional Studies (MS)
Department, Program, or Center
Graduate Programs & Research
Advisor
Khalil Al Hussaeni
Recommended Citation
Amin, Dana, "Predicting Cost Contingency in Retrofit Projects using Machine Learning techniques" (2023). Thesis. Rochester Institute of Technology. Accessed from
https://repository.rit.edu/theses/12216
Campus
RIT Dubai
