Jonsson, Gudbjorn F. and Trefethen, Lloyd N. (1997) A Numerical Analyst Looks at the "Cutoff Phenomenon" in Card Shuffling and Other Markov Chains. Technical Report. Unspecified. (Submitted)

PDF
548kB 
Abstract
Diaconis and others have shown that certain Markov chains exhibit a "cutoff phenomenon" in which, after an initial period of seemingly little progress, convergence to the steady state occurs suddenly. Since Markov chains are just powers of matrices, how can such effects be explained in the language of applied linear algebra? We attempt to do this, focusing on two examples: random walk on a hypercube, which is essentially the same as the problem of Ehrenfest urns, and the celebrated case of riffle shuffling of a deck of cards. As is typical with transient phenomena in matrix processes, the reason for the cutoff is not readily apparent from an examination of eigenvalues or eigenvectors, but it is reflected strongly in pseudosprectra  provided they are measured in the 1norm, not the 2norm. We illustrate and explain the cutoff phenomenon with Matlab computations based in part on a new explicit formula for the entries of the "riffle shuffle matrix", and note that while the normwise cutoff may occur at one point, such as for the riffle shuffle, weak convergence may occur at an equally precise earlier point such as .
Item Type:  Technical Report (Technical Report) 

Subjects:  H  N > Numerical analysis 
Research Groups:  Numerical Analysis Group 
ID Code:  1313 
Deposited By:  Lotti Ekert 
Deposited On:  09 Jun 2011 07:23 
Last Modified:  29 May 2015 19:00 
Repository Staff Only: item control page