Related Posts
Bowl Leader
Community Builder
Hi fishes! The hiring manager has asked me for salary expectations of B6 Data Analyst and my experience is 3.5 yrs with a CTC of 6.65. Considering I have to move to Bangalore l'm thinking of asking 13 or 14.. is it a big ask? If so, what do you think is a good ask? Please help. Thanks in advance :)
Natwest group
More Posts
Anyone working as Patient Recruitment Assistant?
Chief
What is moonlighting?
why cts blacklist any company any idea?
Additional Posts in Data & Analytics Consultants
Mentor
Has anyone else begun to resent data science?
New to Fishbowl?
Download the Fishbowl app to
unlock all discussions on Fishbowl.
unlock all discussions on Fishbowl.
Ours are SQL then python or R
The latter is pseudo code though, more to gauge how you think and do you understand the algos.
I don’t care that you memorized:
Numpy.rf(mydata) blah blah
I care that you know why you chose random forest, why it works, why it might not, and how to explain it to the business
usually a normal modeling task they give you the data and you implement whatever classification or regression model of your choice
both state farm and ibm had similar modeling tests for me
Surprised there’s a coding assessment for an entry level position. There are so many other, more important, skills to test for at entry level. I’d be much more interested in your ability to learn to code than whether you actually can. I’d be a bit wary of this role if I’m honest, sounds like they want the ‘finished article’ at entry level salary.
Entry level DS itself is kind of a misnomer imo - need to be very careful when evaluating DS offers. Half of the teams hiring them don't actually have any plan ahead of time and often don't even have the minimal infra built out to properly support them.
As for the coding assessment, I have yet to run across a DS role that doesn't either have a coding assessment or a set of take home tasks (at least several dozen interviews taken and given over the last decade.) If they're hiring DS without doing any vetting at all I would run the other way.
If it's Python make sure you are comfortable with the basic data summarisation, manipulation and quality checks in pandas.
Look up feature engineering techniques.
Dont confuse classification, regression, time-series problems.
Read up on testing methods, cross validation etc.
And learn the choice of accuracy metrics best for the problem at hand.
For entry-level positions, I usually expect familiarity with how to use joins, aggregation functions, and transformation functions (transpose, pivot, melt) in Python or SQL. Bonus points if you know how to use window functions.
I don’t expect perfect syntax (e.g. don’t expect you to memorise the exact parameters of various pandas functions), but instead evaluate how well you walk through your coding logic and structure the problem. I also evaluate how well you take feedback/suggestions.