Data Science and Domain Specific Knowledge

 Data science and domain expertise

I my statistical courses, we often checked sensitivity of the method to specific distributions of the input data. Simpson’s paradox is one of the problems that might be hard to figure out for a data scientist. It is often hard to be a domain expert in the are of where data comes from. Certain areas of #manufacturing, #biotech, and #engineering are so intricate that machine learning or data science expertise alone might not be enough to avoid some data fallacies.

This is not a new issue. In other areas of the industry, specialists often need to get degrees in other areas. Most people are familiar with the practice of adding an #MBA degree to their resume. I think that data scientists might be increasingly looking into adding domain-specific #certifications or degrees to do their jobs more reliably.

https://mixpanel.com/blog/avoiding-data-fallacies-and-biases-simpsons-paradox-and-the-importance-of-segmenting-data/

#datascience #domainexpertise #domainexpert #machinelearning #bigdata #datascientists #simpson #datafallacy #dataanalytics #datapreparation #datawrangling #clustering #datasegmentation