LFD Book Forum  

Go Back   LFD Book Forum > Course Discussions > Online LFD course > Homework 2

Reply
 
Thread Tools Display Modes
  #11  
Old 01-20-2013, 04:58 AM
alinsoar alinsoar is offline
Member
 
Join Date: Jan 2013
Posts: 17
Default Re: Q1 and Q2

I will re-phrase the problem 2 with my own words :

We have 1000 bins, and we make 100K experiments of extraction of samples of 10 elements each (with replacement). Each bin has inside it 2 symbols: 'head and 'tail.

Are we allowed to apply the Hoeffding inequality in the following situations:

1. all samples are extracted from the same bin

2. each sample is extracted from a randomly bin

3. each sample is extracted from the bin that provided at the Kth experiment the minimal freq. of heads.

Am I right to think the problem so or am I wrong ?

In my case, which is the hypothesis set ?
Reply With Quote
  #12  
Old 01-20-2013, 07:04 AM
palmipede palmipede is offline
Member
 
Join Date: Jan 2013
Posts: 13
Default Re: Q1 and Q2

For Q1, the exact distribution of n_min can be computed and the analytical answer matches my empirical answer but none of the listed answers exactly.
Reply With Quote
  #13  
Old 01-20-2013, 08:09 AM
alinsoar alinsoar is offline
Member
 
Join Date: Jan 2013
Posts: 17
Default Re: Q1 and Q2

Quote:
Originally Posted by palmipede View Post
For Q1, the exact distribution of n_min can be computed and the analytical answer matches my empirical answer but none of the listed answers exactly.

For me it does , read closely : νmin is closest to:
Reply With Quote
  #14  
Old 04-12-2013, 11:57 AM
jlaurentum jlaurentum is offline
Member
 
Join Date: Apr 2013
Location: Venezuela
Posts: 41
Default Re: Q1 and Q2

Hello:

I understand that the mu's are the expected values the relative frequency of heads in N=10 flips. Clearly, mu_1=mu_crand=0.5. Mu_min is different from this value, but it can be calculated analytically. I found that all three sample frequencies (nu's) converge to their respective mu values for all three coins, regardless of whether one coin is fixed and the other 2 sample frequencies are obtained by looking at the 1000*10 flips. This is to be expected, as the law of large numbers states that all sample frequencies converge to the true population frequencies. However, there is no option in question 2 for "all three coins satisfy hoeffding's inequality". Where is my error?
Reply With Quote
  #15  
Old 04-12-2013, 12:52 PM
yaser's Avatar
yaser yaser is offline
Caltech
 
Join Date: Aug 2009
Location: Pasadena, California, USA
Posts: 1,477
Default Re: Q1 and Q2

Quote:
Originally Posted by jlaurentum View Post
Hello:

I understand that the mu's are the expected values the relative frequency of heads in N=10 flips. Clearly, mu_1=mu_crand=0.5. Mu_min is different from this value, but it can be calculated analytically.
Hi,

All the coins are fair by assumption, so all the \mu's are 0.5.
__________________
Where everyone thinks alike, no one thinks very much
Reply With Quote
Reply

Thread Tools
Display Modes

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off

Forum Jump


All times are GMT -7. The time now is 07:27 AM.


Powered by vBulletin® Version 3.8.3
Copyright ©2000 - 2019, Jelsoft Enterprises Ltd.
The contents of this forum are to be used ONLY by readers of the Learning From Data book by Yaser S. Abu-Mostafa, Malik Magdon-Ismail, and Hsuan-Tien Lin, and participants in the Learning From Data MOOC by Yaser S. Abu-Mostafa. No part of these contents is to be communicated or made accessible to ANY other person or entity.