Hi,
In this practice lab, we have feature scaling for all the models except the logistic regression. Do we not need a feature scaling because the logistic regression is not distanced-sensitive?
Nice catch @Andrew_Xu! It is just lucky that that dataset has only two features, and their means and standard deviations are not too much different, and the network is not too deep. However, we need to keep in mind to always scale our features.
Cheers,
Raymond