Multi-objective Multiplexer Decision Making Benchmark Problem

This paper proposes a novel multi-objective decision making benchmark problem. The problem addresses the need in the multi-objective decision making realm for an easy to construct, scalable benchmark problem in the vain of the DTLZ, ZTD, and WFG problems. The problem is inspired by a real-world decision making problem that pilots face in the cockpit. The new problem is an amalgamation of two well-established problems within the literature, the DTLZ and multiplexer problems. The problem additionally makes use of the main concepts and ideas from Robust Decision Making and Multi-scenario Multi-objective Robust Decision Making, especially as these problems enable decision making problems to be somewhat converted into an optimization task. The problem is showcased here and is solved initially using a modified multi-objective optimization variant of a Learning Classifier System, which shows superior results when compared to a random agent.