General question: Is there any situation when Decision trees analysis benefit of standardised input data?In PLS data is standardized by default. For decision tree analysis its not.
Are there any situations when it is recommended to use standardized input data?
Br
// PTR
AnalyticsRe: General question: Is there any situation when Decision trees analysis benefit of standardised input data?Here's my two cents...PLS and all linear modeling methods I can think of standardize the predictor variables because it's generally a good idea for computational and interpretational reasons for things like parameter estimates, standard errors, and such. Since decision trees are a non linear modeling method, my sense is standardization doesn't provide many benefits so it's not generally practiced.
I'd love to read other's comments.
