Interesting slide deck from Josh Wills,Tom Pierce, and Jeff Hammerbacher of Cloudera.
The mention of “Pareto optimization” reminds me of a debate tournament judge who had written his dissertation on that topic. Who carefully pointed out that it wasn’t possible to “know” how close (or far away) a society was from any optimal point. 😉 Oh well, it was a “case” that sounded good to opponents unfamiliar with economic theory at any rate. An example of those “critical evaluation” skills I was talking about a day or so ago.
Not that you can’t benefit from machine learning and Hadoop. You can, but ask careful questions and persist until you are given answers that make sense. To you. With demonstrable results.
In other words, don’t be afraid to ask “stupid” questions and keep on asking them until you are satisfied with the answers. Or hire someone who is willing to play that role.