We introduce a new algorithm for the induction of classifiers from data, based on Bayesian networks. Basically this problem has already been examined from two perspectives: first, the induction of classifiers by learning algorithms for Bayesian networks, second, the induction of classifiers based on the naive Bayesian classifier. Our approach is located between these two perspectives; it eliminates the disadvantages of both while exploiting their advantages. In contrast to recently appeared refinements of the naive Bayes classifier, which captures single correlations in the data, we have developed an approach which captures multiple correlations and furthermore does a trade-off between complexity and accuracy. In this paper we evaluate the implementation of our approach with data sets from the machine learning repository and data sets artificially generated by Bayesian networks.