. Asked about research in my current role, digging into details about diffusion models I worked on (and he clearly did not know any basics and seemed he did not want to admit that). Quite unpleasant. Then technical question about parameter estimation from noisy observations of Gaussian random variables.
Some of mathematical statistical questions.
New York Bike demand, statistics, linear algebra, involving mathematical proof. medium to hard level coding.
linear regression coding test
probability of getting even number of H
What is the relation between regress Y on X and regress X on Y? Be prepared to derive the relation between 2 r-squares analytically.
ML projects; why certain techniques were used; feature engineering
The data analysis interview asks you how you would predict some number. First you go through data ideation (if you could gather any reasonable data, what would you want?), then you talk about how you would go about engineering features and cleaning your data, then you talk about what kind of model you would use and why. They really drill into your answers, and you are likely not going to pass if you don't have a lot of experience in predictive modeling. Watch the mock interview video on their website for an easy example of what this is like.
Do you have any background on statistics?
Spatial data and common models.
Viewing 3451 - 3460 interview questions