Hi Miguel:   Thank you for the immedate reply.   I don't think the data set has something special. It has a lot of descriptive variables of the clients of a credit company and the target variable is constructed on the category the company gives to its clients. The categories goes from 1 to 6 and if you have 3 or more you are considered a bad client so we just build with a CASE that says that the target variable. Every client has a category so there're no missing values.    I'm going to try what you suggest and then I tell you how it goes, now I only have this picture of the window of results of the Interactive Grouping node, I attach the rest later.      Thank you! ... View more