Re:Boston Housing Price Dataset was Removed from scikit-lear...
by josephcsible on Friday December 30, 2022 @05:25PM
(#34190810)
I wonder if that was the overall proportion for the entire surrounding area at the time. If so, then B=0 would mean "this looks like what you'd expect if where people lived had nothing to do with race", and increasing values of B would be increasing evidence to the contrary.
Re:Boston Housing Price Dataset was Removed from scikit-lear...
by HPsquared on Friday December 30, 2022 @05:42PM
(#34190994)
It is kind of strange, I don't understand why they would use that. Here's a plot of the function:
https://www.wolframalpha.com/input?i=plot+%5B1000*%28x-0.63%...
It's seems like a weird choice. If there are 62% or 64% then the feature will yield the same value.
I think it would make more sense just to include percentage of households where at least one member of the household has a certain ethnicity.
I don't think it's a offensive to analyze demographic information in the aggregate. In fact this happens all the time.