How can I apply the equation to make binary classification

[ghost]

I have a question about SymbolicClassifier. The resulting equation is like, x2/x1 + x3- x5. How can I apply the equation to make binary classification ? My goal is to make binary classification on my data.

[IFFranciscoME]

you might consider using the output of the equation as an input to an activation function, a sigmoid for example will therefore create a logistic regression classifier.

[ghost]

Thank you for your response. Since it is a classifier, can not we learn it again to find a threshold values for the output equation based image so that it makes binary classification ? It would be much applicable if the SymbolicClassifier learns to output the equation like it. class: 0 if x2/x1 < 8 and x3- x5 > 9. How can I write such a custom function ......

[trevorstephens]

https://gplearn.readthedocs.io/en/stable/intro.html#classification a sigmoid is used by default to transform the output

[ghost]

yes sigmoid has been used to classify the data. However, my requirement is to get equation from _Program after the sigmoid is applied. What is printing now is before the sigmoid is applied.

[trevorstephens]

Yes, it's possible to add the sigmoid to the representation printed on the screen. I can put this in the long-overdue next release.

[trevorstephens]

I think it makes sense to do it, thanks for the report