Data-Driven Decision-Making in IT Projects: A Regression Application in Software Estimation

IT-related projects are among the most challenging to manage due to their inherent complexity, uncertainty, and the intangibility of their outcomes. However, a data-driven decision-making approach can help tackle these challenges and increase the chance of IT project success. Machine learning techniques can be used to predict uncertain aspects of IT projects, such as the estimation of time, cost, and effort, when applied to historical data collected from previous or similar projects. This approach can allow the project manager to make informed decisions based on facts rather than human judgment or hunch. In this work, we review the potential of data-driven approaches in the management of IT projects, and we demonstrate a use case that applies three regression algorithms, including polynomial regression, linear regression, and partial least squared (PLS), to predict software effort estimation based on the NASA93 public dataset. The best results were achieved using polynomial regression which scored 82% in R<sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2</sup>, 81% in MAPE, 152 in MAE, 276 in RMSE and 76387 in MSE metrics.