ABSTRACT
Defining if the error is less than 0, the state is 1; while if more than 0, the state is 2. There are two states for the prediction value can’t be absolutely identical to the actual value.
2.2 Markov chain prediction
Markov chain is the random process with parameters and discrete states, which definition is:
Assuming the random process is
for arbitrary conditions i0, i1, · · · , in ∈ I , the state space is I ={i0, i1, · · · , in, · · ·}, the time parameter set is N ={1, 2, · · · , n, · · ·},there is
and
so the random process {Xn, n ∈ N } is called Markov chain.