Targeting market opportunity for port growth: a CART-based decision support system

Complex decisions, such as those facing port managers when targeting market opportunities for growth, are multifaceted and follow complex patterns of sequential and iterative decision subsets that are not well captured nor represented by current parametric modelling approaches. Nonetheless, understanding how such decisions are made and the patterns they follow have important implications for policy formulation and execution. This paper uses a classification and regression trees (CART) framework—a non-parametric decision trees modelling approach--and develops a decision support system to assist port managers in making effective decisions about market opportunities they should pursue for growth. The CART-based decision support system reveals the predictive structure and complex patterns that the market opportunity selection process follows. It also advances our knowledge and understanding of the interaction and behaviour of the key decision variables and the decision makers in a port context.