A note on the two-step matrix splitting iteration for computing PageRank
Autor: | Ting-Zhu Huang, Zhao-Li Shen, Chun Wen |
---|---|
Rok vydání: | 2017 |
Předmět: |
Theoretical computer science
Google matrix Preconditioner Applied Mathematics 010103 numerical & computational mathematics 01 natural sciences law.invention 010101 applied mathematics Arnoldi iteration Computational Mathematics PageRank law Power iteration Fixed-point iteration Matrix splitting Convergence (routing) 0101 mathematics Algorithm Mathematics |
Zdroj: | Journal of Computational and Applied Mathematics. 315:87-97 |
ISSN: | 0377-0427 |
Popis: | Computing PageRank plays an important part in determining the importance of Web pages. Based on the classical power method and the inner-outer iteration proposed by Gleich etźal. (2010), Gu etźal. (2015) presented a two-step splitting iteration, i.e., the power-inner-outer (PIO) iteration, for the computation of PageRank. In this paper, we develop a variant of the PIO iteration by applying multi-step power method to combine with the inner-outer iteration. The new method is denoted as the MPIO iteration, its convergence is analyzed in detail. Numerical experiments on several PageRank problems are used to illustrate the effectiveness of the MPIO iteration. |
Databáze: | OpenAIRE |
Externí odkaz: |