classfication relation in features/examples size

veronicaborrell
veronicaborrell New Altair Community Member
edited November 2024 in Altair RapidMiner
I've  found that  to determine the size of the training set for a classifier (decision tree, naive bayes, Knn,etc) it is recommended that you need at least 6 times the number of cases (samples) as features.. but can't find any reference (paper or publication) about that. Are you aware of any reference please?

Tagged: