1. Your train/test event rate is 10% but oot test data event rate is 5%. Why did you build the model when with 1:10 event rate without checking the real time data for the next 6 months, which is 5%? - I have 14 years of exp including 8 years of consulting. This is the worst interview question I ever had. 2. This question was asked 3 times. You are building a prospect model to target clients for house refinances, at what stage are you building the model? Underwriting, funding? - It was a surprise to receive such questions in interview. I told them that we are identifying the prospects for marketing before even they reach out/start refinancing but they kept asking me the same questions. Looks like they just want to prove they know couple of things about mortgage industry but it was stupid. 3. I mentioned about a summarization project build on llama2 for my finance company built using call transcripts. Interviewer accused me of lying as he thinks I can't build model because ppl discuss ssn numbers and the details about other companies. Seriously? You are in a senior position and you don't know what's synthetic data or how customer data is redacted internally etc? 3. They are super rude from the start during the project review. I lost interest after 15 minutes as they are constants demeaning my current company and the projects we do when they don't even understand the basics of DS.
Director Data Science Interview Questions
453 director data science interview questions shared by candidates
Tell me about a project you worked on from start to finish? What's your management style? What's your thoughts on coding style?
Python question: Provide pseudocode for how to read and concatenate datasets from multiple CSV files in a directory.
How would you solve a particular challenge while in a project? lay down steps /tasks
What is central limit theorem and where it is used.
script a python code to compute the standard deviation
Written exam (with markings!) ranging from "derive Bayes linear regression" to "how to change branches in a git repository"
Describe a past project that went poorly.
Pandas Data Wrangling Python Modeling JSON key-value relationships Modeling set-up Intermediate SQL (sub-selects, aggregations)
Como você lida com uma possível insatisfação do time ao resolver tarefas que não são tão nobres/interessantes?
Viewing 321 - 330 interview questions