A probabilistic proof of cutoff in the Metropolis algorithm for the Erd"os-R'enyi random graph

Mathematics – Probability

Scientific paper

Rate now

  [ 0.00 ] – not rated yet Voters 0   Comments 0

Details

Scientific paper

We consider the Metropolis algorithm for the distribution \pi(x) = \theta^{S(x)} (1+\theta)^{-n} on the hypercube X=\{0,1\}^n, where S(x) is the number of ones in x \in \{0,1\}^n and \theta \in (0,1] is a constant. For n = {nu \choose 2} this distribution corresponds to the Erd"os-R\'enyi random graph model on nu vertices, where each edge is present independently with probability \frac{\theta}{1+\theta}. The lazy random walk Metropolis algorithm for this model specifies a Markov chain (X_t) on X that is known to have cutoff at \frac{1}{1+\theta} n \log n with window size n, a result derived by Fourier analyis in Diaconis and Hanlon (1992) and Ross and Xu (1994). In this work we give a new proof of this result that is purely probabilistic. This is done in the hope that probabilistic techniques will be easier to generalize to other, less symmetric distributions \pi. Our proof uses coupling and a projection to a two-dimensional Markov chain X_t \rightarrow (S(X_t),d(X_0,X_t)), where d(X_0,\cdot) is the Hamming distance to the starting state X_0.

No associations

LandOfFree

Say what you really think

Search LandOfFree.com for scientists and scientific papers. Rate them and share your experience with other people.

Rating

A probabilistic proof of cutoff in the Metropolis algorithm for the Erd"os-R'enyi random graph does not yet have a rating. At this time, there are no reviews or comments for this scientific paper.

If you have personal experience with A probabilistic proof of cutoff in the Metropolis algorithm for the Erd"os-R'enyi random graph, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and A probabilistic proof of cutoff in the Metropolis algorithm for the Erd"os-R'enyi random graph will most certainly appreciate the feedback.

Rate now

     

Profile ID: LFWR-SCP-O-210218

  Search
All data on this website is collected from public sources. Our data reflects the most accurate information available at the time of publication.