Confusing Linear Regression Learner Node Description

Hi everyone,

I'm a bit confused by the Node Description of the Linear Regression Learner.

1) It says it "performs a multivariate linear regression". But as far as I know, that means that there are multiple target variables and not just one scalar. But there is only one drop-down menu to select one target variable. Why does it say that the regression is multivariate?

2) The description links to a Wikipedia page containing a section about estimation techniques. Can someone tell me which technique is implemented in this node? Ordinary least squares?

Thank you very much for your help!

Hi Mol1hua,

1) The "multivariate" refers to the features rather than the target. You predict the target based on multiple inputs, as opposed to a bivariate regression where you have one target and one feature.

2) It uses OLS



Dear Roland,

Thank you so much for your answer!

It might be helpful to define "multivariate" in the node description, since it links to a Wikipedia article which states "This term [multiple linear regression] is distinct from multivariate linear regression, where multiple correlated dependent variables are predicted, rather than a single scalar variable."

Your hint about OLS helped me a lot!

Kind regards,