What would you do if a costumer calls and did everything like the protocol said and wants his money back?
Data Scientists Interview Questions
54,225 data scientists interview questions shared by candidates
The data challenge is pretty standard, give you a data set and let you build a predictive model and put together some insights. Phone and onsite interviews are all technical interviews. Some statistics (sampling, hypothesis testing, power calculation, etc.), some product sense (whether to launch some features, why some metric changed, etc.), also a CS-like question.
I was asked about the way I would design the "people you may know" application of LinkedIn and about the way I would group "similar" people and jobs with people etc. More or less it was a discussion about collaborative filtering , classification and prediction models (k-means clustering, bayesian model, decision trees ).
How many colors do you need to paint a graph that is built from a union of two trees?
behavioral questions about location and fit
How would you correlate each device a person visits the website with back to that person?
If you had 1000 factors for a ML model, how would you go about reducing this and how would you know when to stop. reducing
Two random variables have a uniform distribution between 0 and 1. What will be the expected value of the difference of these variables?
What is maximum likelihood estimation?
Imagine you have N pieces of rope in a bucket. You reach in and grab one end-piece, then reach in and grab another end-piece, and tie those two together. What is the expected value of the number of loops in the bucket?
Viewing 201 - 210 interview questions