dabl: Automatic Machine Learning with a Human in the Loop |SciPy 2020| Andreas Mueller

แชร์
ฝัง
  • เผยแพร่เมื่อ 10 พ.ย. 2024

ความคิดเห็น • 4

  • @Tech_Datasavvy
    @Tech_Datasavvy 4 ปีที่แล้ว +1

    Hey Andy ...dabal.clean should not remove missed value column, it might be a different pattern . It can tell missing value columns list and treat it separately. Other steps like constants and sparse categorical or ID info is fine . Some case Foreign key can be used for counts too as a feature. Looks messy if pandas profiling can be used with dabl package in junction to get a beautiful DF with all the details to be used or not used will help...what’s your view.

    • @AndreasMueller
      @AndreasMueller 3 ปีที่แล้ว

      Missing values are not actually removed, they are encoded as a separate category. If you have ideas of how to integrate better with pandas profiling, I'm all ears!

  • @MatthewRocklin
    @MatthewRocklin 4 ปีที่แล้ว +1

    Fun talk! Thanks Andy.