I don't quite understand the generalized additive model behind the GAM package in R. It seems quite powerful with the ability to easily find complex relationships and confidence intervals for these as seen in the R Graphical Manual. Are there any big limitations to these models and is this why I cannot find an implementation in sklearn for Python?
Contents
hide
Best Answer
As mentioned in the comments, a propensity to overfit is a limitation of GAMs. Another limitation is that the model will lose predictability when the smoothed variables have values outside of the range of training dataset. Essentially, you are sacrificing predictability outside of your data range for precision within your data range.
Similar Posts:
- Solved – Limitations to generalized additive model (GAM)
- Solved – Limitations to generalized additive model (GAM)
- Solved – Is additive logistic regression equivalent to boosted decision stumps
- Solved – Multiple regression and OLS. How to choose the best “non-linear” specification
- Solved – why different results with summary()and anova() function in linear mixed effect model