LFD Book Forum  

Go Back   LFD Book Forum > Book Feedback - Learning From Data > Chapter 1 - The Learning Problem

Thread Tools Display Modes
Old 08-05-2014, 10:23 AM
yusunchina yusunchina is offline
Junior Member
Join Date: Aug 2014
Posts: 6
Default Question on the Selection of Samples from the Bin

In the multiple hypothesis section, we are presented the analogy of having let's say m bins, and a random sample of marbles of size N from each bin. My question is, are those m random samples the same marbles in each bin (only probably colored differently), or are they different for each bin (for example, randomly selected when we sample each bin)?

To clarify: suppose bins 1...m each contains marbles 1...M (M >> N). Let's say N = 4, and in the sample for bin 1 we select marbles # 2, 3, 4, 5 and color them. Then for bin 2...m, do we still sample marbles # 2, 3, 4, 5 or do we randomly select 4 probably different numbers for each bin?

I tried to explain this myself, here are two contradictory explanations:
1. From the coin example give, we are randomly sampling for each bin, since the corresponding marbles in the bins are colored the same but the selected samples are colored differently.
2. From the training perspective, they are the same samples. If a bin corresponds to the entire population and a sample is the training set from the population, for selection of each hypothesis we are using the same training set (not generating new ones from the population).

So which view is correct? Thank you very much!!!!
Reply With Quote
Old 08-10-2014, 06:57 PM
yaser's Avatar
yaser yaser is offline
Join Date: Aug 2009
Location: Pasadena, California, USA
Posts: 1,478
Default Re: Question on the Selection of Samples from the Bin

Your thoughts are correct. In the case that corresponds to multiple hypotheses and a given training set, the marbles are the same among different bins. This does not affect the calculation because (1) Per bin, the marbles are still picked independently of each other, and (2) Jointly, we apply the union bound that covers all possible dependencies between the different bins.
Where everyone thinks alike, no one thinks very much
Reply With Quote
Old 08-11-2014, 03:49 PM
yusunchina yusunchina is offline
Junior Member
Join Date: Aug 2014
Posts: 6
Default Re: Question on the Selection of Samples from the Bin

Thank you very much Professor! It makes me feel much safer to have this confirmation from you. It's very generous of you to spend time answering questions from us readers!
Reply With Quote

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 07:18 PM.

Powered by vBulletin® Version 3.8.3
Copyright ©2000 - 2021, 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.