It is correct to apply methods as McNemar test or 5x2CV paired t test in order to compare two classification models with high accuracy? For example, Logistic regression and KNN in the iris dataset. I'm not sure if it is correct to apply this statistical methods because bot of them have a very high accuracy, near to 100% or 100% in some cases, so the binomial distribution (percentage of correct classified examples) approach not a normal distribution but a truncated normal. Thanks for the help.



Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge that you have read and understand our privacy policy and code of conduct.

Browse other questions tagged or ask your own question.