Loading [MathJax]/extensions/TeX/boldsymbol.js

 

 

 

Understanding the basics

We can always expand \boldsymbol{w}(0) in terms of the right eigenvectors \boldsymbol{v} of \boldsymbol{W} as

\begin{equation*} \boldsymbol{w}(0) = \sum_i\alpha_i\boldsymbol{v}_i, \end{equation*}

resulting in

\begin{equation*} \boldsymbol{w}(t) = \boldsymbol{W}^t\boldsymbol{w}(0)=\boldsymbol{W}^t\sum_i\alpha_i\boldsymbol{v}_i= \sum_i\lambda_i^t\alpha_i\boldsymbol{v}_i, \end{equation*}

with \lambda_i the i^{\mathrm{th}} eigenvalue corresponding to the eigenvector \boldsymbol{v}_i .

If we assume that \lambda_0 is the largest eigenvector we see that in the limit t\rightarrow \infty , \boldsymbol{w}(t) becomes proportional to the corresponding eigenvector \boldsymbol{v}_0 . This is our steady state or final distribution.