AHP-Based Determination of Warning Grade in a Warranty Claims

Two perspectives on developing better decision capabilities for a warranty system can be identified: one involving the inclusion of a ‘learning’ module and the other the inclusion of a ‘prioritization’ capability. This paper demonstrates how a warning process can be included in a warranty system by coupling with a neural network’s learning capabilities. In addition to the neural network, a method is employed for assigning priorities to warning criteria by using the analytic hierarchy process (AHP). Thus, it is possible to construct an integrated system with three components: the warranty system, the AHP module, and the neural network system. A case study is provided to enhance the accuracy of warning/detection judgment in a warranty system for automobile companies, having many factors related to the warranty system.