Project Portfolio Reliability: A Bayesian Approach for LeAgile Projects

Abstract We propo se an applied Bayesian learning approach for continuous planning and evolution of information system projects and portfolios. Unlike traditional project management approaches for information system, the proposed approach considers the cumulative effect of all past experiences to achieve continuous performance and reliability prediction. The results of quantitative comparisons with other common estimation approaches, such as non-learning point estimates and traditional Bayesian approach, using real case data indicate that the proposed approach can generate a more realistic metric to continuously plan and measure the performance of evolving LeAgile projects or portfolios. This study can support decision makers, engineering teams, and management by supplying a practical and scalable project performance prediction tool for continuous planning and system evolution.