Quote:
Originally Posted by yaser
Hoeffding cannot be directly applied, but more general forms that use expected values and variances of general random variables (rather than just the probability of an event and the frequency of occurence of that event) can be applied.

I think I'm starting to converge on an understanding. So the basic form of Hoeffding introduced in lecture only addresses estimating the probability of an event. I checked the Wikipedia entry for the Hoeffding inequality (
http://en.wikipedia.org/wiki/Hoeffding%27s_inequality) and it states a more general form that covers the expected value of bounded random variables. Is this one of the more general forms that you mentioned? Looks like the Hoeffding equation in the lecture could be derived from this form by setting the lower and upper bounds to 0 and 1. Am I on the right track here?
Quote:
Originally Posted by yaser
Hoeffding can be directly applied to (1), the event being "ten heads out of 10 flips," whereas (2) involves the expected value of a random variable as discussed above.

Could the basic form of Hoeffding be used for (2) where the event is described as flipping a head on a single coin toss?