Random Waks on Graphs An Overview图上的随机游动的一个概述.pptVIP

* Power iteration Power Iteration is an algorithm for computing the stationary distribution. Start with any distribution x0 Keep computing xt+1 = xtP Stop when xt+1 and xt are almost the same. * Power iteration Why should this work? Write x0 as a linear combination of the left eigenvectors {v0, v1, … , vn-1} of P Remember that v0 is the stationary distribution. x0 = c0v0 + c1v1 + c2v2 + … + cn-1vn-1 * Power iteration Why should this work? Write x0 as a linear combination of the left eigenvectors {v0, v1, … , vn-1} of P Remember that v0 is the stationary distribution. x0 = c0v0 + c1v1 + c2v2 + … + cn-1vn-1 c0 = 1 . WHY? (see next slide) * Convergence Issues1 Lets look at the vectors x for t=1,2,… Write x0 as a linear combination of the eigenvectors of P x0 = c0v0 + c1v1 + c2v2 + … + cn-1vn-1 c0 = 1 . WHY? Remember that 1is the right eigenvector of P with eigenvalue 1, since P is stochastic. i.e. P*1T = 1T. Hence vi1T = 0 if i≠0. 1 = x*1T = c0v0*1T = c0 . Since v0 and x0 are both distributions * Power iteration v0 v1 v2 ……. vn-1 1 c1 c2 cn-1 * Power iteration v0 v1 v2 ……. vn-1 σ0 σ1c1 σ2c2 σn-1cn-1 * Power iteration v0 v1 v2 ……. vn-1 σ02 σ12c1 σ22c2 σn-12cn-1 * Power iteration v0 v1 v2 ……. vn-1 σ0t σ1t c1 σ2t c2 σn-1t cn-1 * Power iteration v0 v1 v2 ……. vn-1 1 σ1t c1 σ2t c2 σn-1t cn-1 σ0 = 1 σ1 ≥…≥ σn * Power iteration v0 v1 v2 ……. vn-1 1 0 0 0 σ0 = 1 σ1 ≥…≥ σn * Convergence Issues Formally ||x0Pt – v0|| ≤ |λ|t λ is the eigenvalue with second largest magnitude The smaller the second largest eigenvalue (in magnitude), the faster the mixing. For λ1 there exists an unique stationary distribution, namely the first left eigenvector of the transition matrix. * Page