07.08.2013 Views

beamer - Vrije Universiteit Amsterdam

beamer - Vrije Universiteit Amsterdam

beamer - Vrije Universiteit Amsterdam

SHOW MORE
SHOW LESS

You also want an ePaper? Increase the reach of your titles

YUMPU automatically turns print PDFs into web optimized ePapers that Google loves.

Probabilistic Averages IV<br />

.<br />

Corollary<br />

.<br />

Suppose that the Markov chain satisfies the unichain condition with a finite<br />

set of transient states (|T| < ∞). Then for all states i, j ∈ I<br />

1<br />

lim<br />

n→∞ n<br />

. (i.e., independent of the initial state).<br />

n∑<br />

k=1<br />

p (k)<br />

ij<br />

= πj<br />

Proof: πj = 0 for transient j; and fij = 1 for recurrent j (see slide 7). Then<br />

apply slide 15.<br />

c⃝ Ad Ridder (VU) SOR– Fall 2012 17 / 36

Hooray! Your file is uploaded and ready to be published.

Saved successfully!

Ooh no, something went wrong!