from Hacker News

The Rise of the Citizen Data Scientist [pdf]

by altairiumblue on 2/20/19, 9:22 AM with 3 comments

• by tenkabuto on 2/20/19, 4:36 PM

  > Citizen Data Scientist: A person who creates or generates models that leverage predictive or prescriptive analytics but whose primary job function or education is outside of the field of statistics and analytics. Their formal job title might be something like data analyst, business analyst, or similar.

  Data analysts' "primary job function or education is outside of the field of statistics and analytics"? Nevertheless, I think that this might be an interesting shift in job roles. However, I'm worried that users of such tools would unwittingly biasing the models that they create.

• by altairiumblue on 2/20/19, 5:10 PM

  Personal opinion - I really dislike the term and all the overpriced drag-and-drop, "automated", "easy-to-use" tools that are associated with it. Generally, I would choose open-source over licensed software and a programming language over a GUI.

  It would be interesting to hear the thoughts of other people who've used similar solutions as well as Python/R for the same process.

• by laiden_swallow on 2/20/19, 6:48 PM

  Machine and deep learning tools are reaching higher levels of abstraction all the time. Libraries like fast-ai make them accessible to anyone with a basic understanding of Python.

  I don't think the advent of the 'Citizen Data Scientist' is even about automation. It's about providing an interface that allows application/experimentation without a deep understanding of the underlying statistics and architectures.

  I think it's totally reasonably to expect that, within the next couple of years, there will be state of the art deep learning techniques that don't require any more technical expertise to apply than operating a smart phone.