+ Reply to Thread
Results 1 to 3 of 3

Thread: Problem with bayesian updating example

  1. #1
    Points: 1,269, Level: 19
    Level completed: 69%, Points required for next Level: 31

    Posts
    2
    Thanks
    0
    Thanked 0 Times in 0 Posts

    Problem with bayesian updating example




    Hi everybody,

    I have an example in bayesian statistics which I've worked on for some hours now, without significant success. The problem is the following (all of this is structured and prepared in the attached Excel sheet):

    There are two outcomes (W and L) and three signals for these outcomes (B, H, S). The following is known:
    P(B)=0.6; P(H)=0.3, P(S)=0.1, P(W|B)=0.7; P(W|H)=0.25, P(W|S)=0.5.

    Now I would like to calculate the other probabilities (i.e. P(W), P(B|W), etc.) - which I think I have managed to do. Furthermore, I want to calculate the updated probabilities after I have observed one signal of B, which I have so far failed to succeed at.

    I have attached an Excel sheet that contains the data and my results so far.

    Thanks for your help!

    Cheers,
    Bayesianski.
    Attached Files

  2. #2
    TS Contributor
    Points: 22,410, Level: 93
    Level completed: 6%, Points required for next Level: 940

    Posts
    3,020
    Thanks
    12
    Thanked 565 Times in 537 Posts

    Re: Problem with bayesian updating example

    Since \{B, H, S\} forms a partition, we have

    P(W) = P(W|B)P(B) + P(W|H)P(H) + P(W|S)P(S)

    And from the Bayes Theorem:

    P(B|W) = \frac {P(W|B)P(B)} {P(W)}

  3. #3
    Points: 1,269, Level: 19
    Level completed: 69%, Points required for next Level: 31

    Posts
    2
    Thanks
    0
    Thanked 0 Times in 0 Posts

    Re: Problem with bayesian updating example


    Hi,

    Thanks for your reply! I also have:

    P(L) = 1 - P(W),

    P(L|B) = 1 - P(W|B), P(L|H) = 1 - P(W|H), P(L|S) = 1 - P(W|S),

    P(B|L) = P(L|B)P(B) / P(L).

    But now how do I update all of these probabilitites after having observed one or two independent draws of B?

    Cheers,
    Bayesianski

+ Reply to Thread

           




Similar Threads

  1. Bayesian kills me
    By SadieKhan in forum Probability
    Replies: 11
    Last Post: 09-23-2010, 12:54 AM
  2. Bayesian Statistics
    By WeeG in forum Statistics
    Replies: 0
    Last Post: 04-10-2009, 05:06 AM
  3. bayesian
    By bres6486 in forum Probability
    Replies: 5
    Last Post: 01-16-2009, 12:52 AM
  4. Bayesian or not Bayesian....
    By Marione in forum Statistics
    Replies: 2
    Last Post: 04-26-2008, 03:28 AM
  5. Regarding Bayesian Inference
    By abcd_1641 in forum Psychology Statistics
    Replies: 2
    Last Post: 01-30-2006, 12:23 AM

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts






Advertise on Talk Stats