Premium
Technical details and programming guide for a general two‐way positive matrix factorization algorithm
Author(s) -
Lu Jianhang,
Wu Laosheng
Publication year - 2004
Publication title -
journal of chemometrics
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.47
H-Index - 92
eISSN - 1099-128X
pISSN - 0886-9383
DOI - 10.1002/cem.894
Subject(s) - computer science , implementation , algorithm , documentation , matlab , factorization , programming language
In the past decade a factor analysis model named positive matrix factorization (PMF) has been increasingly used in analysis of environmental data. The PMF computing program is commercially available from its developer, but the technical details of the algorithm and program codes remain unpublished so far. For better documentation and facilitation of its application we present the technical details of a general two‐way PMF algorithm and provide a detailed guide for its programming. We explicitly explain the iterative procedure and construction of matrices and vectors used in the PMF algorithm and illustrate their implementations with ready‐to‐use Matlab codes. We also give our experiences and suggestions on various aspects of PMF programming, such as reducing memory need, improving computational efficiency and adjusting model parameters. The contents presented in this paper are helpful for one who would like to create his/her own PMF computing codes and thus be able to tailor the model for specific tasks or purposes. Copyright © 2005 John Wiley & Sons, Ltd.