Unification of artificial intelligence with optimization for decision support: UNIK approach

Abstract

This paper reviews the UNIK project which unifies the artificial intelligence (AI) with optimization models for decision support. Our goal is to unify the specification of problems to encompass as many hybrid type of problems as possible from both AI and optimization. Currently, the type of problems selected from the AI side are the rules, compatibility constraints, and neural networks, while linear programming and integer programming models from the optimization side. The key issue to be resolved for the unification is the unified representation of optimization model and rules so that they can understand each other. The second issue is designing the unified architecture and devising the solution method. This paper reviews the above issues one by one with illustrative applications