Use of Data science and machine learning algorithms in order to develop solutions for the management of various capital-intensive projects is explained through one of its applications in the estimation of project owner cost. Owner cost prediction using machine learning techniques help in project cost allocation at the portfolio level. The application of ML algorithms using key factors like project size, project location, duration, extension, turned out to improve the accuracy of the cost estimation by 14%. The final Mean absolute percentage error of the model turned out to be 25.16%. This research is a foundation for application in other areas of project management. It will also serve as a reference for application of machine learning models using historical data to predict the project owner cost for various capital-intensive projects.