How to create stochastic matrix for these probability values?

• Aug 29th 2011, 01:27 PM
How to create stochastic matrix for these probability values?
I hope I am posting in the correct forum.

If the weather is good today, there is a 20% chance it will be fair and 7% chance it will be bad tomorrow.
If the weather is fair today, there is an 18% chance it will be good and a 22% chance it will be bad tomorrow.
If the weather is bad today, there is a 20% chance it will be good and a 30% chance it will be fair tomorrow.

(a) How to construct and label stochastic matrix describing the weather trends?
(b) What is the chance of bad weather on Thursday if on Monday there is a 88% chance of good weather and a 2% chance of bad weather.
(c) Is there a steady state vector if there is an equally likely chance of good, bad, or fair weather today?

Attempt:
As we can see
P(fair tomorrow/good today) = 0.20
P(good tomorrow/fair today) = 0.18

How to make the stochastic matrix? Can we do as follows?
1st column = good today
2nd column = fair today
1st row = good tomorow
2nd row = fair tomorrow

Therefore, the matrix will be
? ......... 0.18 ......... 0.20

.20 ...... ? ............. 0.30

.07 ...... 0.22 .......... ?

We can find the missing values assuming that the sum of entries in each column = 1
Is this correct? If no, then how to write the correct stochastic matrix?

And how to answer the other parts?
Thank you.
• Aug 29th 2011, 04:54 PM