locked
Newbie question on Probability - Wet Grass/Sprinkler/Rain Model RRS feed

  • Question

  • Hi,

    I've just started looking at Probability Models, so excuse the newbie question but it seems a fundamental one to me.

    In the Wet Grass/Sprinkler/Rain Model, the CPT values for the Rain node are 0.8 and 0.2 when Cloudy = true.

    Now I get that the probability of rain is higher when it is cloudy, but where does the value of 0.8 come from? Why not 0.75?

    Thanks.

    Monday, October 15, 2012 9:00 AM

Answers

  • It is a natural question to ask. If you look at the documentation for this example, you will see that there are three modes of operation, and these modes of operation are all covered in the running example. Only in the first mode do we assume that we know the CPT values exactly - these are just a given in the problem. In the second mode of operation we do not assume knowledge of these - rather we infer these from observations.

    John

    • Marked as answer by Trossachs Monday, October 15, 2012 8:47 PM
    Monday, October 15, 2012 3:40 PM
    Owner