There were a few questions, but my favorite by far was the expected number of neighboring points inside an epsilon neighborhood. It had the makings of a good challenge without being totally inaccessible. Lots of fun to run through with the interviewer.
IBM Attrition Prediction case study (Kaggle)
What are the biggest challenges you will be faced with in this job, and how will you overcome them?
What is the most common misconception that people have of you?
Hardest problem you faced, transformer-based questions, deep learning vs tradML
Topics: NLP , Deep Learning, MLOps
Tell me about a 'feather in the cap moment'
Explain forecasting methods, deep learning methods eg LSTM
Q: Basic Stats and Project deep dive Q: ML case study like if you are taking over someone's model what are the steps you would follow and follow up questions on that Q: Problem Framing is more like datasets would be given and you have to do easy data manipulation using pandas. Q: Coding: Make sure you study stack, queue, and basic data structures. It would be easy
-Standard ML questions in the online interview - Take home assignment consisted of EDA and training a classifier. - The onsite interviews were standard, the team was great and they treated me very well!
Viewing 1291 - 1300 interview questions