I am new to GAM and pyGAM. I read documentation carefully but need help with someone experienced in pyGAM for some questions.
I have a dataset with around 40 features. My problem is classification and I am planning to use LogisticGAM.
1) I think I have not fully understand how 'splines' used in GAM? Does the model fits a spline for each and every feature?
2) What are the differences between factor term, spline term, tensor term and why do we need these terms?
3) When I read and searched about GAM I understand that some of my features out of 40 may have linear relation while some others have non-linear relationship with target variable. Am I supposed to enable this relationship to model manually for all these features as follows?(I don't think so but don't know what to do) And what should I consider while determining which term(factor, spline, ..) to use?
I am new to GAM and pyGAM. I read documentation carefully but need help with someone experienced in pyGAM for some questions.
I have a dataset with around 40 features. My problem is classification and I am planning to use LogisticGAM.
1) I think I have not fully understand how 'splines' used in GAM? Does the model fits a spline for each and every feature?
2) What are the differences between factor term, spline term, tensor term and why do we need these terms?
3) When I read and searched about GAM I understand that some of my features out of 40 may have linear relation while some others have non-linear relationship with target variable. Am I supposed to enable this relationship to model manually for all these features as follows?(I don't think so but don't know what to do) And what should I consider while determining which term(factor, spline, ..) to use?
gam = LogisticGAM(f(0) + s(1) + ... + s(40))