Lecture 15: Tail bounds
Posted by James on February 15, 2012
Hear are the slides. Reading assignment: BT, Section 5.1.
This entry was posted on February 15, 2012 at 1:34 am and is filed under Lecture, Slides. You can follow any responses to this entry through the RSS 2.0 feed. You can leave a response, or trackback from your own site.
Xin said
I’m wondering is it possible that we can have a version of slides without those giant portraits of mathematicians removed. It seems they always cover a few important stuff.
Thanks.
James Lee said
If you pay attention, you see that they are not covering up anything. The slide with the mathematician is the same as the slide before (minus the mathematician).
Xin said
ahhh. thanks james.:)
Nat said
In the second equation defining chernoff bounds, should that denominator in the exponential be 2, rather than 3? If not, why? These equations don’t make a whole lot of sense to me…
James Lee said
The “3” is correct. Unfortunately, we don’t have time to go over the proof of the Chernoff bound in class. It’s based on the “method of Laplace transforms.” The bounds for the upper and lower tail are not symmetric because a Bin(n,p) random variable is not symmetry around its expectation. They come from an approximation to the true tail behavior. You can see all this discussed on wikipedia: http://en.wikipedia.org/wiki/Chernoff_bound