The Smoothed Analysis of Algorithms
This slides for this talk are not yet finished.
For your reference, I include
this preview of my talk.
It will go though a lot more editing, and
include material on
K-means and learning decision trees.
As the NIPS community is well-positioned to work on Smoothed Analysis,
I'll also throw in a bunch of conjectures, and explain how to go about
proving results in the area.