Dismiss Notice
Join Physics Forums Today!
The friendliest, high quality science and math community on the planet! Everyone who loves science is here!

Not preferenced weight function

  1. Oct 17, 2005 #1
    Dear people of this forum

    Let us suppose we have infinite functions of the form

    F= w1 X1 + w2X2+ w3 X3 + w4 X4+............
    Where w1,w2,w3, w4 are variable weights and X1, X2, X3, X4 are fixed temperatures.

    Now, I have to choose one of the above infinite functions with the requirement that this selected function should be fair, honest, in the sense that it is not preferenced to one side or another . ¿What weigths should I use, w1=w2=w3=w4 ?
     
  2. jcsd
  3. Oct 17, 2005 #2

    mathman

    User Avatar
    Science Advisor
    Gold Member

    If you have an infinite set of x's, you cannot have equal weights. For a finite set you can.
     
  4. Oct 17, 2005 #3
    Would you please, explain me why it is so?
     
  5. Oct 18, 2005 #4

    EnumaElish

    User Avatar
    Science Advisor
    Homework Helper

    What do you mean by "infinite functions"? Is it

    [tex]\lim_{n\rightarrow\infty}\sum_{k=1}^nw_kX_k[/tex]

    ?
     
  6. Oct 18, 2005 #5

    mathman

    User Avatar
    Science Advisor
    Gold Member

    Since the weights must sum to 1, if you have n weights, equal means all 1/n. Try to do it with infinite number - you can't. You will have an infinite number of zeros.
     
  7. Oct 24, 2005 #6
    In relation to my question, If I use equal weights , do this means that, statiscally, we are using a distibution which is not oriented to one side or another, or that it is not biased?
     
  8. Oct 25, 2005 #7

    HallsofIvy

    User Avatar
    Staff Emeritus
    Science Advisor

    Yes, in order that the distribution not be biased toward one outcome, all weights must be equal. As was pointed out before, that is not possible with an infinite number of possible outcomes.
     
Know someone interested in this topic? Share this thread via Reddit, Google+, Twitter, or Facebook

Have something to add?



Similar Discussions: Not preferenced weight function
  1. Weighted Covariance? (Replies: 2)

  2. Negative Weights? (Replies: 9)

Loading...