Good, non technical article in DSC about building predictive models. No hype, suitable for executive use. All this is fairly standard stuff, basic approaches, but I like to see it stated in different ways. You are always trying to be predictive when doing analytics. You are trying to do something predictively better.
Its just how much the system can provide prescriptive results. Less prescription means the analyst/scientist needs to evaluate, compare and implement some alternative results. That's not always easy, and needs its own form of business knowledge, science and art. And sometimes, even often, you can get valuable insight from just visualizing data.
Subscribe to:
Post Comments (Atom)
No comments:
Post a Comment