In the middle of deploying the model, team of data scientists realize that the predictions are "somewhat-off". Troubleshooting on the horizon and what to do. Session will guide you through most common mistakes data scientists and statisticians are making when preparing and engineering the data using T-SQL or any other database system.

Further more, we will explore common statistical and data science mistakes when modeling data, extracting know-how from the data, finding the hidden patterns and running different test against the structural models using mainly R, Python, or Spark. What not-to-do will be replaced with what to-do explanations using sample datasets and sample codes.

Some statistical knowledge or background is a plus!

Presented by Tomaž Kaštrun at SQLBits 2020