A Two-Stage Intervened Decision System With State-Dependent Random Inspection Mechanisms

This paper develops the performability and cost-benefit models for a multi-unit two-stage intervened decision-making system with majority voting rule. The decision process of the system can be divided into two stages, which is an inspection stage (stage 1) and a result submission stage (stage 2). In addition, the system is assumed to have binary inputs and outputs. During the inspection stage, there are two possible states for each decision unit in the system. Each decision unit will be visited at most twice during this stage by the supervisor, and the result for the visit depends on the state of the decision unit. The first visit will be conducted for certain, but the behavior of the second visit will be determined by the result of the first visit. The result provided by each decision unit can be submitted any time during the second stage and there is a checkpoint during this stage to make the process finish earlier. The performance of each decision unit in this stage will be determined by its state at the end of the first stage. The final decision of the system will be made at the end of the second stage. The proposed system can be applied in various applications in real life such as financial investment, product releasing and recall, paper submission and review, and credit evaluation and loan application.