Closed ArnovanHilten closed 1 year ago
Adds the possibility to start from a one-hot encoding as an input. This avoids the linear model assumptions, the effect of 2 is twice the effect of an input with value 1.
Adds the possibility to start from a one-hot encoding as an input. This avoids the linear model assumptions, the effect of 2 is twice the effect of an input with value 1.