Cleverson Alessandra Veronez,  Silvia Modesto Nassar,  Carlos Becker Westphall. 

At every moment the computer network administrator, which has his/her decisions supported basically on data, comes across situations that demand treatment of uncertainty. In spite of the offered services, the computer network management platforms are not capable of identifying problems and suggesting corrective actions, and as a consequence leaves to the administrator the burden of interpreting graphs and variable values. Within this context, ihis work presents a study on the application of probabilistic reasoning, through the implementation a bayesian belief network, which is capable of recognizing the relationships among the figures that represent traffic of the network. In addition, the bayesian network may also be used as a dynamic baseline for proactive expert systems. The outcome of this research resulted in a mechanism, which aims ai supporting the administrator in the decision making process.

http://www.lbd.dcc.ufmg.br:8080/colecoes/sbrc/2000/003.pdf

Processando consulta.