Next: Irreducible and Aperiodic Markov
Up: Ergodicity and Stationarity
Previous: Basic Definitions and Quasi-positive
  Contents
Estimates for the Rate of Convergence;
Perron-Frobenius-Theorem
- Example
(Weather Forecast)
- Let
and
- In Section 2.2.1 we showed that
- the
-step transition matrix
is given by
- and thus
- Consequently
 |
(44) |
where
and hence
if
.
- Remarks
In general geometric estimates of the form (44) for
the rate of convergence can be derived by means of the following
so-called Perron-Frobenius theorem for quasi-positive
matrices.
Theorem 2.6
- Let
be a quasi-positive
matrix with
eigenvalues
such that
.
- Then the following holds:
- (a)
- The eigenvalue
is real and positive.
- (b)
-
for all
,
- (c)
- The right and left eigenvectors
and
of
are uniquely determined up to a constant factor and can
be chosen such that all components of
and
are positive.
A proof of Theorem 2.6 can be found in
Chapter 1 of E. Seneta (1981) Non-Negative Matrices and
Markov Chains, Springer, New York.
- Proof
-
- As
is a stochastic matrix
- obviously
and (41) implies
.
- Thus
is an eigenvalue of
and
and
are
right and left eigenvectors of this eigenvalue, respectively.
- Let now be
- Theorem 2.6 now implies
for
.
Corollary 2.3 yields the following geometric
convergence estimate.
Corollary 2.4

Let

be a quasi-positive transition matrix such that all
eigenvalues

of

are pairwise
distinct. Then
 |
(45) |
- Proof
-
- Example
(Reaching a Consensus)
see C. Hesse (2003) Angewandte Wahrscheinlichkeitstheorie.
Vieweg, Braunschweig, p. 349
- Remarks
-
- For large
the algebraic solution of the linear equation
system (41) can be difficult.
- In this case the estimates for the rate of convergence in
(47) become relevant for the practical implementation
of the method to reach a consensus described in (47).
- We consider the following numerical example.
- Let
and
 |
(49) |
- The entries of this stochastic matrix imply that the third expert
has a particularly high reputation among his colleagues.
- The solution
of the corresponding
linear equation system (41) is given by
i.e. the projection
of the third expert with
the outstanding reputation is most influential.
- The eigenvalues of the transition matrix given in
(49) are
,
and
.
- The ,,basis'' in the rate of convergence given by
(43) is
whereas Corollary 2.4 yields the following
substantially improved geometric rate of convergence
where
denotes the second largest eigenvalue of the
stochastic matrix
given by (49).
Next: Irreducible and Aperiodic Markov
Up: Ergodicity and Stationarity
Previous: Basic Definitions and Quasi-positive
  Contents
Ursa Pantle
2006-07-20