3 ms·
Kids with an interest in data science are going to read this and find it baffling, and the references won't help much. There's nothing baffling about it.
by anigbrowl 4y ago
Kids with an interest in data science are going to read this and find it baffling, and the references won't help much.
There's nothing baffling about it.
- civilized 4y agoThe cited Medium post says many things. Most of them seem to be background or diversions. I did my best to pick out the essential points and discuss them in my comment above. To the extent I could understand, those essential points seemed questionable. For what it's worth, another commenter posted this and I found it much clearer, if still imperfect: https://fairlearn.org/main/user_guide/datasets/boston_housing_data.html https://fairlearn.org/main/user_guide/datasets/boston_housin... The FairLearn authors here explain, like I did above, that the ethical issues depend on use case, and are not inherent to the data or the dataset developers' choice of transformation for the B variable.