A general model of a system which moves from one state to state is described and applied to concrete problem. It is shown that such systems tend to a steady-state eventually.
The transition probability is the probability that if the system is in state at any one observation, it will be in state at the next observation.
A transition matrix is any square matrix with nonnegative entries, all of whose column sums are one.
The probability vectors (column vectors of a transition matrix) for are said to be the state vectors of a Markov process if the component of is the probability that the system is in the state at the observation.
If is the transition matrix of a Markov process and is the state vector at the observation, then
A car rental agency has three rental locations, 1, 2, and 3. A customer may rent a car from any of the three locations and return the car to any of the three locations. The manager finds that the customers return the cars to the various locations according to the following probabilities:
where stands for the probability of renting a car from location and return it to location Suppose a car is initially rented from location number 2.
(1) Find the state vector
(2) Predict .
We define , and
which is designed to compute
So all state vectors are equal to to three decimal places.
What if we set ? We obtain that
A transition matrix is regular if some integer power of it has all positive entries.
If is a regular transition matrix, then as
where the are positive numbers such that
If is a regular transition matrix and is any probability vector, then as
where is a fixed probability vector independent of .
Note that if is regular, then as then for some Thus which is a fixed vector and we set it to be
be an transition matrix of a Markov process. State vector
is called a stable state or steady state of the Markov process if
The transition matrix . Find the steady-state vector .
Method 1: We compute as we did in the previous example.
Method 2: If , then , which is equivalent to solve a homogeneous linear system. (We build Identity matrix with Scientific Workplace by using ''Matrices + Fill Matrix + Identity''.) We set as follows:
. We solve
and the "Solution is :