Markov Chains with Rare Transitions and Simulated Annealing.
Abstract
This document considers Markov chains in which the entries of the one-step transition probability matrix are known to be of different orders of magnitude and whose structure (that is, the orders of magnitude of the transition probabilities) does not change with time. For such Markov chains the author presents a method for generating order of magnitude estimates for the t-step transition probabilities, for any t. He than notices the algorithms of the simulated annealing type may be represented by a Markov chain which is approximately stationary over fairly long time intervals. Using the results he obtains a characterization of the convergent cooling schedules for the general class of algorithms of the simulated annealing type.
Document Details
- Document Type
- Technical Report
- Publication Date
- Sep 01, 1985
- Accession Number
- ADA161598
Entities
People
- John N. Tsitsiklis
Organizations
- Massachusetts Institute of Technology