R&D project portfolio selection in public organizations

Portfolio selection is one of the main tasks performed in organizations that allocate resources to research and development in various areas. In general, the problem of R&D project portfolio selection (RDPPS) is to choose a set of project proposals that optimize certain impact measures designated by the decision maker. The proposals are characterized by one or more social objectives and geographic regions within a certain area of influence. Typically a project is broken down into tasks to which objectives, goals and resources requirements are linked. In this paper, we propose a Multiobjective Mixed Integer Lineal Mathematical Model that allow the representation of R&D projects by its tasks, partial allocation of resources to tasks, and interdependencies among tasks or projects. Also some experimental results for very large scale instances are presented and dicussed.