Note on the Error Analysis of the Generalized Rankannihilation Method
作者:
M. Nicolaas (Klaas) Faber,
期刊:
Analytical Letters
(Taylor Available online 1999)
卷期:
Volume 32,
issue 14
页码: 2899-2906
ISSN:0003-2719
年代: 1999
DOI:10.1080/00032719908543014
出版商: Taylor & Francis Group
关键词: Bilinear calibration;GRAM;Bias;Variance;Chemometrics
数据来源: Taylor
摘要:
The generalized rank annihilation method (GRAM) is a method for curveresolution and quantitation that uses two data matrices simultaneously, i.e., onefor the unknown and one for the calibration sample. Requirements have beenderived that ensure the unique resolution of the analyte of interest in thepresence of unknown interferences if the data matrices are free fromexperimental error. In this paper, it is shown that the same requirements allowfor correct determination of bias and variance in the quantitative results obtainedby GRAM if the data matrices are not free from experimental error.
点击下载:
PDF (263KB)
返 回