Towards a multi-agent system (MAS) based Credit Reference Bureau

Credit risk forms an enormous source of losses for money lending institutions. Mechanisms and models to implement early warning credit risk systems have been put in place to aid business risk managers make well-informed decisions. With the establishment of credit reference bureaus (CRB), it´s imperative to have a real-time mechanism of collecting data, analyzing it and reporting the knowledge in it. Predictive machine learning techniques can learn the pattern behavior and classification thereof of the imminent credit risk that potential clients pose to the financial institutions. Credit reference bureaus are central bank regulated institutions that monitor the population´s credit response as experienced by financial institutions. Banks and micro-financing institutions form credit reference bureaus´ clientele. Intelligent multi-agents systems presents an opportunity to make credit risk decision making more timely, efficient and less human-centric.