I'm trying to use stacking when predicting for the infamous Iris dataset. Also, I'd like to build to stacked classifier by myself which means I don't want to use mlxtend because it's too "easy" and out of the box. I'm using python 3 and working on windows 10.
I'm not having any issue choosing algorithms or during cross-validation but when I try to stack the base learners ( gaussian and multinomial naive bayes, logistic regression) which gave me a string answer : the species of iris ; with the meta-learner (gradient boosting classifier), I can't seem to have a string as an answer, only a probability and it my accuracy score is much lower than any of the individual base learner.
Could anyone direct me at some code and/or explanations as to what I should be doing instead?
Thanks a lot!