Methodology and Computing in Applied Probability

, Volume 12, Issue 4, pp 763–773 | Cite as

The Power of Choice in the Construction of Recursive Trees

Article

Abstract

The power of choice is known to change the character of random structures and produce desirable optimization effects. We discuss generalizations of random recursive trees, grown under the choice to meet optimization criteria. Specifically, we discuss the random k-minimal (k-maximal) label recursive tree, where a set of k candidate parents, instead of one as in the usual recursive tree, is selected and the node with minimal (maximal) label among them is assigned as parent for the next node. These models are proposed as alternatives for D’Souza et al. (Eur Phys J B59:535–543, 2007) minimal and maximal depth models. The advantage of the label models is that they are tractable and at the same time provide approximations and bounds for the depth models. For the depth of nodes in label models we give the average behavior and exact distributions involving Stirling’s numbers and derive Gaussian limit laws.

Keywords

Random structure Random tree Gaussian law Stirling number Power of choice 

AMS 2000 Subject Classifications

Primary 60C05 60F05 05A05 05C05 

Preview

Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.

References

  1. Bergeron F, Flajolet P, Salvy B (1992) Varieties of increasing trees. In: Lecture notes in computer science, vol 581. Springer, Berlin, pp 24–48Google Scholar
  2. D’Souza R, Krapivsky P, Moore C (2007) The power of choice in growing trees. Eur Phys J B59:535–543MATHCrossRefMathSciNetGoogle Scholar
  3. Mahmoud H (1991) Limiting distributions for path lengths in recursive trees. Probab Eng Inf Sci 5:53–59MATHCrossRefMathSciNetGoogle Scholar
  4. Smythe R, Mahmoud H (1994) A survey of recursive trees. Theorya Imovirnosty ta Matemika Statystika 51:1–29 (in Ukrainian); an English translation appears in Theory of Probability and Its Applications 1996MATHGoogle Scholar

Copyright information

© Springer Science+Business Media, LLC 2009

Authors and Affiliations

  1. 1.Department of StatisticsThe George Washington UniversityWashingtonUSA

Personalised recommendations