A multi-agent system architecture for statistics managing and soccer forecasting

Prediction markets have substantially grown during the last years. In particular, sports forecasting is an important field of application. In this paper we present a new forecasting system oriented to sports. This proposal is a multi-agent system conformed by several intelligent and independent agents. The agents perform different complementary tasks. This heterogeneous approach provides a powerful collaboration environment that increases the accuracy of the predictions. We present the results of our experiments and show the promising preliminary results.