Hello,
I am hoping that someone can help me trouble shoot an error message I am getting on Homework 3, Question #8. I successfully ran the linear regression model and made the binary Pass3 variable, but when I try to run the support vector machine node on the partitioned data, I get the attached error message. Any suggestions on how to trouble shoot it?
Thanks,
Teresa
Teresa. I see that when modeler is having problems with the data. My guess it that Pass3 has been made so that it is all "0" or "1", I would check that first. The second possibility is that the training dataset set is too small or that the Pass3 is all "0" or "1" there. Basically, this error seems like it is about the data and not about SVM. Best, Dave
Hi Teresa,
Check the roles of variables for your linear and svm model. They should be different for each model. I use a set of role for my linear model and but I run svm, I add another type node to change the roles of those variables according to the instruction in the homework.
