I think the way I understand item a) of this question may not be the same as the teachers intended. I consider the “range” as the maximum minus minimum distance in the data, and not necessarily the StandardScaler will make the ranges of each feature similar. In the case of the idealized examples, yes, because the “density” of points is similar, but if any of those features has an outlier, it will not necessarily be transformed to a range similar to the other features. Maybe the teachers thought of the “range” as a more flexible term? I would like the opinions of others
1 Like
I understand your point here. I think that we should refer to the previous dataset such that one avoids thinking about valid corner cases (such as stated).