When analyzing sensory data, one option is to give equal weight to each sensory panelist, the other -- to only consider the best panelist whose responses correlate highly with the panel mean. Between these two extremes, where does your approach to data analysis fall, and why?
Do you think your SVM is overfitting on the data and that is why you get good training results but bad results on actual data?
Code example with fiverr data of sellers and buyers. I asked to give the distribution of number f sellers with 1 service for sail, 2 services for sail, etc.
What is your favorite ML algorithm?
Do you know about Regression and Classification problems? *If you are interviewing for senior data science position, is this your level of question?
What is your career goal?
How do you evaluate the model performance? Which metric will you pick?
What do you bring to the role and where do you see your career headed?
I was asked to code a left join in Python.
Programming questions about Python. Asked about processing circular reads.
Viewing 721 - 730 interview questions