"Point #1: it’s about the model. Point #2: the model is contingent on exploration of the data and bringing in information from external sources. Which begs the question, if you’re not doing this, are you doing “data science”? I happen to think the answer is “no”.Be wary of “data scientists” who… well, aren’t | SecurityCurve
Meaning, if you hire a bunch of consultants to come in and “work their magic” on your data set, that is a non-starter if you accept the original premise by which the term (and some might argue, the discipline) came about. Namely, because you’re ignoring the “complex” part — the “salient part of the analysis”. Can you do that? Sure. Is it a good idea? I don’t think it is."
Tuesday, April 09, 2013
Be wary of “data scientists” who… well, aren’t | SecurityCurve
Excerpt from a timely data science reality check