Login | Register

Algorithms for random ranking generation


Algorithms for random ranking generation

Xu, Liqun (2000) Algorithms for random ranking generation. Other thesis, Concordia University.

[thumbnail of MQ54338.pdf]
Text (application/pdf)


Given a ranking of size n , most of the existing ranking models have relatively small numbers of parameters (around n , or less). Having a small number of parameters do help facilitate the application of a ranking model. But on the other hand, it restricts the capacity for the model to describe the innate structure of a ranking population. In this report, we suggest a random ranking generator with ( n - 1) 2 parameters. The increased number of parameters enables the generator to simulate ranking populations with greater flexibility. We also suggest the use of a n x n probability matrix (P-matrix) as a device for specifying the targeted ranking population. In the P-matrix, each cell is the probability of an item being assigned a certain rank. We provide an algorithm that estimates, from a given P-matrix, the parameters for the generator. Numerical examples show that using the P-matrix based parameter estimation algorithm, the proposed generator provides better simulation to the targeted rank data.

Divisions:Concordia University > Gina Cody School of Engineering and Computer Science > Computer Science and Software Engineering
Item Type:Thesis (Other)
Authors:Xu, Liqun
Pagination:vi, 47 leaves ; 29 cm.
Institution:Concordia University
Degree Name:Major reports (M.Comp.Sc.)
Program:Computer Science and Software Engineering
Thesis Supervisor(s):Lam, Clement
Identification Number:QA 76 M26+ 2000 no.7
ID Code:1221
Deposited By: Concordia University Library
Deposited On:27 Aug 2009 17:17
Last Modified:13 Jul 2020 19:48
Related URLs:
All items in Spectrum are protected by copyright, with all rights reserved. The use of items is governed by Spectrum's terms of access.

Repository Staff Only: item control page

Downloads per month over past year

Research related to the current document (at the CORE website)
- Research related to the current document (at the CORE website)
Back to top Back to top