Projects i have worked on, how I handle certain issues, what technologies I know. Some basic questions during the L2 round
Staff Data Scientist Interview Questions
118 staff data scientist interview questions shared by candidates
Why LR could give you negative orediction
Build a simple marketing mix model. What are your recommendations about how to allocate marketing budget across channels based on your model?
One thing to note: I wasn't asked a single question on statistics or machine learning.
What are L2, L1 regularization? What is gradient descent? Explain transformers.
Technical screen involved mainly SQL questions as this was for a data modeling focused role. Some on SQL logic (it's open ended, so you can answer with whatever dialect you're familiar with) and some on writing performant queries.
The recruiter just asked some high-level questions about my past experience, and insisted that they were handing out take-home assignments to everyone when clearly it was false.
How would you do this statistical analysis task, calling it "AI". However, it was mentioned every Staff Data Scientist has left because there is no role definition and they still aren't sure what the objective of the role. You will be given a take home dataset of 1000 records and are only allowed 3 hours of working time including create a presentation for 15 minutes of interviews. In my case, I used libraries I had built to do RL hyper parameter optimization with XGBoost to use PDP plots to understand which variables impacted the target variable the most. Software engineers were the interviewers that only knew XKCD memes and didn't understand basic ML.
Tell me about your model deployment strategies.
Tell me about your experience. Go over past projects and say where you failed and where you succeeded.
Viewing 101 - 110 interview questions