The Move-To-Front Rule for Self-Organizing Lists with Markov Dependent Requests
We consider the move-to-front self-organizing linear search heuristic where the sequence of record requests is a Markov chain. Formulas are derived for the transition probabilities and stationary distribution of the permutation chain. The spectral structure of the chain is presented explicitly. Bounds on the discrepancy from stationarity for the permutation chain are computed in terms of the corresponding discrepancy for the request chain, both for separation and for total variation distance.
Key wordsMarkov chains self-organizing search move-to-front rule convergence to stationarity separation total variation distance coupling
AMS(MOS) subject classifications
Unable to display preview. Download preview PDF.
- Diaconis, P. (1993). Notes on the weighted rearrangement process. Unpublished manuscript.Google Scholar
- Dobrow, R. P. (1994). Markov chain analysis of some self-organizing schemes for lists and trees. Ph.D. dissertation, Department of Mathematical Sciences, The Johns Hopkins University.Google Scholar
- Fill, J. A. (1995). An exact formula for the move-to-front rule for self-organizing lists. J. Theo. Prob.,to appear.Google Scholar
- Hendricks, W. J. (1989). Self-organizing Markov Chains. MITRE Corp., McLean, Va.Google Scholar
- Knuth, D. (1973). The Art of Computer Programming, Searching and Sorting, Vol. III. Addison-Wesley, Reading, Mass.Google Scholar
- Phatarfod R. M., and Dyte, D. (1993). The linear search problem with Markov dependent requests. Preprint.Google Scholar
- Rodrigues, E.R. (1993) The performance of the move-to-front scheme under some particular forms of Markov requests. Preprint.Google Scholar