This paper addresses the problem of selecting a robust portfolio of projects in the context of limited resources, multiple criteria, different project interactions and several types of uncertainty. A portfolio of projects is considered an undoubtedly robust choice if for a given uncertainty domain that affects the costs and/or the benefits of the projects there is no other portfolio that does not cost more and simultaneously may provide more overall benefit. We present a new decision support system, PROBE (Portfolio Robustness Evaluation), and the algorithms it implements. PROBE identifies all efficient portfolios and depicts the respective Pareto frontier within a given portfolio cost range, and permits users to analyze, in depth, the robustness of selecting a proposed portfolio. The robustness evaluation starts by identifying competitor portfolios to the proposed portfolio, its similarities and differences in project composition to its competitors, and the regret a decision-maker may have by selecting the proposed portfolio instead of a competitor.
- restricted efficiency
- portfolio decision analysis
- resource allocation
- portfolio robustness
- multicriteria decision support system