Want to Truly Master Scikit-Learn? 2 Essential Tips from Core Developer Himself


TDS interviews Andreas Muller from Scikit-learn who shares 3 practical machine learning techniques all data scientists needs to know.



Comment List

  • Towards Data Science
    November 14, 2020

    0:42 What is your professional background?
    1:21 Your popular ML book has no math or formulas. Why?
    2:22 How did you get involved with Scikit-learn as a Core Developer?
    3:51 How does the open-source community work for Scikit-learn (workflow and ownership)?
    5:43 As a former release manager of the package, what did you do?
    7:18 What are common mistakes/inefficiencies you've seen among those implementing Scikit-learn?
    9:02 What is pipelining?
    10:29 What other metrics other than accuracy do you recommend for machine learning?
    11:49 Tools or techniques in Scikit-learn that you feel are being underutilized or undervalued?
    13:05 How do you feel about Catboost, LightGBM, Pytorch, and other players in the growing ML library field?
    16:58 How did your philosophy of democratizing AI come about?
    18:07 How do you tackle imbalanced data? What are strengths/weaknesses of two main techniques?
    21:08 Real problems of SMOTE in practice.
    23:50 Words if wisdom for the those interested in getting into open-sourcing.

  • Towards Data Science
    November 14, 2020

    Amazing! Thanks a lot for the gold content 😀 +1 subscription

  • Towards Data Science
    November 14, 2020


Write a comment