Learning of Differential and Integral Calculushas been a challenge for students of the technologicalareas in Brazil. This paper presents a study on attitudesand behaviors of students within and outside theclassroom that interfere with its performance in thatdiscipline in an Engineering course. For this purpose aclosed questionnaire was applied for a group of students.The analysis of questionnaires was based on BayesianNetworks, using the process of KDD (KnowledgeDiscovery Date) through the shell Netic. Based onconditional probabilities, the results revealed severalfactors of success or failure related to the performance ofstudents in the discipline. With these results can bepresented to students at the beginning of the semester,concrete data that reinforce the attitudes and behaviorsfor successful adoption in Differential and IntegralCalculus in an Engineering course.Index Terms ⎯ Bayesian Networks. Differential andIntegral Calculus. Knowledge Discovery Data. Learning.