Multi-criteria portfolio construction for research and innovation policy; A heuristic optimisation approach

Portfolio approaches are increasingly used in research and innovation policy to balance multiple objectives when allocating resources across competing opportunities. In practice, however, constructing such portfolios is challenging because the number of feasible combinations grows rapidly with the size of the candidate set. This paper proposes a practical method for constructing multi-criteria portfolios when only a limited number of items can be selected. The approach formulates portfolio construction as a compromise optimisation problem that identifies the feasible portfolio closest to a reference point defined by the best attainable performance on each criterion in normalised criterion space. To address the resulting combinatorial problem, the paper introduces a computationally efficient heuristic search procedure. The method enables the systematic construction of balanced portfolios, improves transparency relative to ranking-based approaches, and provides information on the robustness of portfolio choices. The framework is flexible and can be extended to incorporate weighted criteria, budget constraints, and logical dependencies between items.