Suppose I have fit a classification model through Genetic Programming. The output is a symbolic expression that is a function of the covariates, and that associates to each data point a class, such that, if the value of the function at that point is larger than a threshold, it is classified as a positive example; if smaller, classified as a 0 or negative example.

My question is, can we interpret the value of the discriminant function as the probability that the data point belongs to each class?

Say, if a data point has a discriminant Value of 0.98, can we say that the probability the model assigns to the data being of class 1 is 98%, and conversely that it is of class 0 with a probability of 2%?


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.