Beyond Big Data | Matthew Salganik | TEDxPrincetonU

[zp]

Way cool! This really opens up options for doing things at a different level.

[williamwolfe]

Love the example and it was explained well. The idea of using sparse "truth" data to train the net, and then using that to extrapolate to the whole population is powerful, yet limited, because it is still guess work, as was stated. And, the poverty maps did not match completely, and seemed to be off the mark in several regions -- that should have been addressed. Maybe the difference is inconsequential, as implied by the speaker, but it looked significant.

[backtoemocovers]

I learned about Mathew because of a book called Noise.

[tiffanyylee]

he kind of looks like john krasinski from the side sometimes

[江兆峰]

Although the presentation wasn't skillful, we can still extract the information behind. To SOCIAL SCIENTIST!