Thread by @FlorisgoesF

Preparing for a technical interview for a #DataScience position? These are some of the questions that typically allow me as an interviewer to quickly distinguish between juniors and mediors, including some quick tips 🧵. #Python #pythonprogramming #DataScientist #Jobs

All questions about SQL. Not the hardest thing to learn, but many #DataScientists only start to learn the value of SQL when they actually become part of a dev team. I’m not only talking about SELECT * FROM table, but also about joins, truncates, partitions and constraints.

Interacting with an API. Make sure you know your requests (GET, POST, PUT, DELETE, PATCH), as well as the #Python requests library.

Inner joins vs outer joins. Simple question? Maybe. But when you put #DataScientists on the spot in an interview, feeling pressured to give good answers, more often then not they struggle to answer if they are not joining dataframes on a daily basis.

How would you evaluate {model x} for {use-case y}. Every #DataScientist can tell you about RMSEs, ROC AUC and accuracy, fewer can tell you about log loss and Brier Score, and even fewer manage to properly explain why they chose a given KPI, for a given model, in a given use-case

Clustering vs. Classification, and I don’t mean “what’s the difference”. Given a use-case where you want to differentiate between n groups, how would you approach it and what ML type would you use and why. Example: train a model to distinguish between tactical styles in #soccer

Accuracy or Explainability. There are no wrong answers here, but it’s all about the why. Obviously #DataScientists would prefer both, but what if that’s not possible?

These might seem simple questions you learn all about in uni. But I’m not looking for text book answers here. Questions will be wrapped in a practical use-case, and I’m looking for an applied answer. More difficult than you think.

One might say that some of these (SQL, APIs etc) are not core #DataScience. But I like to ask about more than ML, stats and feature engineering, as those skills are essential when working in a development team. Productizing #DataScience is more than #Python #pythonprogramming

Prepare yourself, not only by knowing the text-book answers. Set-up a local database, connect to a public API with #OpenData and write some #Python code to collect, process and store data. Then try to answers some research questions using the data. You just build a pipeline.