Author： Husson François Pagès Jérôme
Publisher： Taylor & Francis Ltd
ISSN： 1360-0532
Source： Journal of Applied Statistics, Vol.32, Iss.4, 2005-06, pp. : 341-350
Disclaimer: Any content in publications that violate the sovereignty, the constitution or regulations of the PRC is not accepted or approved by CNPIEC.
Abstract
We often want to complete the interpretation of the usual graphs (x, y) with additional quantitative variables. The Prefmap method (vectorial model) proposes a representation of these additional variables but this representation has some drawbacks when the variables x and y are correlated. To solve this problem, we propose to substitute the coefficients of the linear regression by the coefficient of the PLS regression in the Prefmap method. The graph obtained is made operational thanks to contour lines of quality of representation and it becomes richer than the Prefmap one.
Related content
Access to resources
Recommend
Optimum covariate designs in split-plot and strip-plot design set-ups
By Dutta Ganesh
Journal of Applied Statistics, Vol. 36, Iss. 8, 2009-08 ,pp. :