Model-based analysis lost the big data race, because it's easier to apply stats to data than it is to model underlying causal behaviors
The consequence of this is that we have fundamentally basic models that adapt only to what features we, with limit understanding, point to.
-
-
This is in some sense fair; in many cases, we don't have the requisite mathematics to fully explain the behaviors we observe.
-
But it is terrifying when we elide their existence entirely. Instead, we use simplified techniques to understand very complex phenomena...
-
and then use these limited conclusions to make actual policy decisions. In the worst case, these conclusions determine who to drone-strike.
-
But we chug ahead because computing power is cheap, databases are ultimately pretty easy, and the ramifications ultimately seem distant.
-
This is not to say that data science is irrelevant. But there is a missing component and that is a posteriori validation of models...
-
along with an explicit process in which to do so. Data analysis is a step in a solution process, not a solution in and of itself.
-
We too often confuse provability in a mathematical context with truth in the broad strokes of human experience.
End of conversation
New conversation -
Loading seems to be taking a while.
Twitter may be over capacity or experiencing a momentary hiccup. Try again or visit Twitter Status for more information.