Project Development and Evaluation of Project Risk Analysis Model Using Machine Learning Algorithms

Abstract Projects are undertaken by the contemporary organizations to experience significant growth, deal with industry competition and achieve sustainability. The increase in size, complexity and duration of the projects resulted in facing several risks and this project is undertaken to develop a project risk analysis model with the help of machine learning algorithms. Identifying that machine learning is feasible for making risk analysis process autonomous, random forest and logistic regression were the two machine learning algorithms utilized to develop the model and then train it. Evaluation of the project risk analysis model showed its effectiveness in recording high number of true positives and true negatives during the detection activity of project risks. From evaluation, random forest is identified as the effective algorithm due to high prediction accuracy, robustness and avoiding overfitting problem. Keywords: Project risks, risk analysis, supervised machine learning, random forest