Machine Learning Intern Interview Questions

8,210 machine learning intern interview questions shared by candidates

You have a partially observable environment with evolving dynamics (non-stationary transition and reward distributions). Logged data comes from multiple behavior policies. How would you estimate the expected return of a new policy and safely improve it, without deploying it, while accounting for uncertainty in both the dynamics and the behavior policies?
avatar

Senior Machine Learning Scientist

Interviewed at Grafton Sciences

5
Oct 17, 2025

You have a partially observable environment with evolving dynamics (non-stationary transition and reward distributions). Logged data comes from multiple behavior policies. How would you estimate the expected return of a new policy and safely improve it, without deploying it, while accounting for uncertainty in both the dynamics and the behavior policies?

OA: One longest interval with sum limit the other is string parsing and hashmap. Probability technical: the classical Seattle raining problem, uniform sphere sampling from uniform distribution, expected number of times of two consecutive coin heads, logistic regression gradient descent from scratch coding behavioral: talking about previous projects, what if disagree with supervisor, asking for help or solve by oneself.....
Jan 11, 2022

OA: One longest interval with sum limit the other is string parsing and hashmap. Probability technical: the classical Seattle raining problem, uniform sphere sampling from uniform distribution, expected number of times of two consecutive coin heads, logistic regression gradient descent from scratch coding behavioral: talking about previous projects, what if disagree with supervisor, asking for help or solve by oneself.....

Viewing 5931 - 5940 interview questions

Glassdoor has 8,210 interview questions and reports from Machine learning intern interviews. Prepare for your interview. Get hired. Love your job.