I try to create a classification tree. My dependent variable is participation which is coded as a categorical 1/0 variable. Participation = 1 means a person works, participation = 0 means a person doesnt work. For a better interpretation i set participation = 1 as the positive class. The majority class is the positve one.
Now I read in some papers that tells me the precision and recall are important measures to evaluate the performance of a classifier. But in all the papers the minority class was the positive one. In my work the negative class is the minority class. Should i know change negative class to the positive one in my confusion matrix?
I think if i do this my results become a bit confusing to interpret, because participation = 1 would be a negative class 0. How can i solve the problem to get a good interpretation for my recall and precision if the minority class is the negative one participation = 0 ?