Created by Knut M. Synstadfrom the Noun Project

Stationary

Definition (Stationary)

Given a Stochastic Process Xn,n0X_n,n\ge0. XnX_n is a time homogeneous (or time-invariant) Markov chain if it is a Markov chain, and there exists {pij:i,jS}\{p_{ij}:i,j\in S\} such that for any i,jSi,j\in S, P(Xn+1=jXn=i)=pij,\mboxforanyn0P(X_{n+1}=j|X_n=i)=p_{ij}, \mbox{ for any } n\ge0 where pijp_{ij} is the probability of transitioning into state jj when in state ii or if its conditional pmfs pXiXi1p_{X_{i}|X_{i-1}} do not depend on time index ii: P(Xi=bXi1=a)=P(X2=bX1=a) i2, a,bXP(X_i=b|X_{i-1}=a)=P(X_2=b|X_{1}=a)\ \forall i\ge2, \ \forall a,b\in\mathcal{X}