Enviar rexistro por email: Learning Parsimonious Classification Rules from Gene Expression Data Using Bayesian Networks with Local Structure