LFD Book Forum (http://book.caltech.edu/bookforum/index.php)
-   Homework 1 (http://book.caltech.edu/bookforum/forumdisplay.php?f=130)

 11lara11 10-10-2013 05:56 AM

Homework1 Q4 and Q5 - Please show workings

I have attempted Q4 using the following:
The epsilon here is close to 0.55 becasue v=0 and u=0.55. The N=10.

P[|v-u|>0.55] < 2e^-2(0.55^2)10
................... < 0.00471

but the solution options don't come anywhere close to this. What am I doing wrong? I think I am making a similar mistake in Q5 becasue I get 4.715 for that becasue it is simply the solution above *1000 becasue M=1000.

 yaser 10-10-2013 11:49 AM

Re: Homework1 Q4 and Q5 - Please show workings

Quote:
 Originally Posted by 11lara11 (Post 11561) I have attempted Q4 using the following: The epsilon here is close to 0.55 becasue v=0 and u=0.55. The N=10. P[|v-u|>0.55] < 2e^-2(0.55^2)10 ................... < 0.00471 but the solution options don't come anywhere close to this. What am I doing wrong? I think I am making a similar mistake in Q5 becasue I get 4.715 for that becasue it is simply the solution above *1000 becasue M=1000.
The course is currently is session at edX, and there are active discussions in their forum there, so you may want to take a look.

 iamds 06-09-2015 05:40 AM

Re: Homework1 Q4 and Q5 - Please show workings

I am also having this problem in the question , please help whats been done wrong in this .
Thank you :confused:u

 yaser 06-19-2015 01:31 AM

Re: Homework1 Q4 and Q5 - Please show workings

The problems ask for the probability, not a bound (Hoeffding or otherwise) on the probability. Try using independence to calculate the probability exactly. If you get Problem 4 right, it is fairly straightforward to get Problem 5, too using simple rules of combining probabilities of events.

 All times are GMT -7. The time now is 11:40 AM.