A hybrid system for modeling credit ratings for US airports

Developing credit ratings for public institutions requires procedures that enable analysts to consider a complex combination of quantitative and qualitative data in a manner that is verifiable, consistent, and reasonably transparent. The procedures must also provide rationale for the ratings and the ability to monitor those ratings on a regular basis. Rating agencies have tried to achieve these objectives by using expert analysts and rating committees. However, they find themselves increasingly challenged about the quality of their work, particularly the ongoing monitoring of all previously rated institutions. This work describes a successful hybrid system that combines fuzzy logic and expert systems to model the rating process for public airports.