Quantum-inspired ensemble approach to multi-attributed and multi-agent decision-making

Abstract

Decision-making is the process of selecting a belief among several other candidate beliefs. The selected belief is usually regarded as a decision while the agent that performs the decision-making process is called as a decision-maker. The decision-making process that makes use of multiple attribute cue pattern as well as multiple decision makers to select a belief is called as multi-attribute and multi-agent decision making. This research proposes quantum inspired approach to multi-attribute and multi-agent decision making process considering the ability of quantum theories to model several complex human cognitive processes as reported in the literature. The main contributions of this work are: a) quantum object based encoding of multi-attribute cue pattern b) ensemble based representation of multi-agent decision making. Further, an analogy between proposed work and double slit experiment to illustrate the interference effect in decision making is also presented. The novelty of this work lies in: a) proposing a quantum inspired ensemble approach to model multi-attribute and multi-agent decision making b) the proposed work outperforms the quantum approach as well as ensemble approach to the decision making with Pima Indiana Diabetes (PID) dataset. The proposed work yields a best accuracy of 90.5% with PID dataset.