Data science

New
cloutch

New

cloutch
Apr 25, 2019 17 Comments

Do you have to be good in math to do data science

comments

Want to comment? LOG IN or SIGN UP
TOP 17 Comments
  • Google
    tsukino

    Go to company page Google

    BIO
    w
    tsukino
    No - undergrad level linear algebra, calculus, probability are sufficient
    Apr 25, 2019 5
    • Google
      tsukino

      Go to company page Google

      BIO
      w
      tsukino
      i certainly felt that most graduate math courses progressed much quicker than the undergrad ones
      Apr 26, 2019
    • New / Eng
      abuhr3i

      New Eng

      abuhr3i
      Also ML was a grad level course used by DS
      Apr 26, 2019
  • New
    ==<>===

    New

    ==<>===
    Yes
    Apr 25, 2019 0
  • New
    DATAxx2033

    New

    DATAxx2033
    A lot of Data Scientists are mathematicians before the title of a data scientist. Even if you aren’t good at math, you CAN learn to be great it’s just a matter of having the drive to take on the math that’s needed to become a strong data scientist. I have my bachelors in CS and (I’m back in school now, and will be pursuing my phD in the future) you’re definitely going to have to understand the concepts of certain math. I hated math, always did, cursed it to h** in HS but my love for data, computers and research never stopped so I pursued my degree and faced the math. So you will have to get on board with math if you want to knock it out of the park in interviews. I always say if the subject intimidates me then it must be right for me.

    Some questions they might ask on an interview:

    “1) Compare and contrast the mathematical machinery used in boosting vs. bagging. How does this impact algorithm speed and accuracy?

    2) Derive a likelihood ratio test based on these two models/statistical distributions.

    3) Give a distance metric and ask candidate to derive a nonparametric statistical test.

    4) What was the last machine learning paper you read? Critique the method and suggest potential ways to improve that algorithm (speed or accuracy).

    5) Create your own neural network model using a novel mapping function.

    6) How can topology be leveraged to extend statistical methods? How have TDA tools like persistent homology interacted with statistics?”
    Oct 11, 2019 0
  • New
    max10chrs

    New

    max10chrs
    Well, 'math' is broad and 'good' is relative. So I don't know how to answer. But yes, data science requires understanding of statistics.
    Apr 25, 2019 1
    • Affirm
      hong_y

      Go to company page Affirm

      hong_y
      Right. Good compared to the average 1st year math PhD? No. Good compared to the average sophomore (across all majors) in a top-100 US undergrad school? Yes.
      Dec 28, 2019
  • New
    mlearning

    New

    mlearning
    Yes. At least the basic fundamentals.
    Math is a very big topic
    Apr 25, 2019 0