|COOKIES: By using this website you agree that we can place Google Analytics Cookies on your device for performance monitoring.|
On the mixing time of random conjugacy walks
If you have a question about this talk, please contact Julia Blackwell.
Let G be a finite graph and consider a random walk on this graph. How long does it take for this walk to be well mixed, i.e., to be close to its equilibrium distribution? A striking phenomenon, discovered in the early 80’s by Aldous and Diaconis independently, is that convergence to equilibrium often occurs abruptly: this is known as the cutoff phenomenon. In this talk we shall consider the classical example of random transpositions over the symmetric group. In this case, Diaconis and Shahshahani used representation theory to prove that such a cutoff occurs at time (1/2) n log n. We present a new, probabilistic proof of this result, which extends readily to other walks where the step distribution is uniform over a given conjugacy class. This proves a conjecture of Roichman (1996) that the mixing time of this process is (1/C) n log n, where C is the size of the conjugacy class. This is joint work with Oded Schramm and Ofer Zeitouni
This talk is part of the Probability series.
This talk is included in these lists:
Note that ex-directory lists are not shown.
Other listsSir Richard Stone Annual Lecture Statistics of Prof Philip Dawid Science talks
Other talksDefining Autism Since 1979: The Sciences of Social Impairment Creating a Shallow-Water Wave Environment Density Embedding Methods: Quality Control and Efficiency for Large Metal Clusters Rogue states: altered dimensionality of neural circuit activity in Fragile-X mice Prison psychiatry, with particular focus on learning disability Designing drugs at the 5HT1B serotonin receptor: from Migraine to Pulmonary Hypertension