Constraint and Optimization Techniques for Supporting Policy Making

Public institutions develop policies and plans in order to achieve economic and social development while preserving the environment. This is a difficult task where computational intelligence data analysis techniques can provide an important contribution. The policy maker has to take decisions by optimizing a set of often conflicting objectives and satisfying a set of constraints. The aim is to reduce negative impacts and enhance positive impacts of plan decisions on the environment, society and economy, exploiting all the data that is available on the territory that is targeted. Up to now, only agent-based simulation models have been proposed in the literature for policy making. In these models, agents represent the parties involved in the decision making and implementation process and simulation is used in order to evaluate the impacts of the policy. Agent-based simulation models provide “individual level models”: we claim that the policy planning activity needs also a global perspective that faces the problem at a global level while tightly interacting with the individual level model. We thus propose a mathematical optimization model that can be applied to regional planning. In the model, decision variables represent political decisions (for instance the magnitude of a given activity in the regional plan), potential outcomes are associated with each decision by considering the available data, constraints limit possible combination of assignments of decision variables, and objectives can be used either to evaluate alternative solutions, or translated into additional constraints. The model has been solved with Constraint Programming techniques. The model has been tested on the Emilia-Romagna regional energy plan. The results have been validated with an expert in policy making and impact assessment to evaluate the accuracy of the results