Model population analysis in chemometrics

• Published in: Chemometrics and intelligent laboratory systems Vol. 149; pp. 166 - 176
• Main Authors: Deng, Bai-Chuan, Yun, Yong-Huan, Liang, Yi-Zeng
• Format: Journal Article
• Language: English
• Published: Elsevier B.V 15.12.2015
• Subjects: Applicability domain; Chemometrics; Model evaluation; Model population analysis; Outlier detection; Variable selection; Model evaluation; Outlier detection; Model population analysis; Applicability domain; Chemometrics; Variable selection
• ISSN: 0169-7439; 1873-3239
• DOI: 10.1016/j.chemolab.2015.08.018

Model population analysis (MPA) is a general framework for designing new types of chemometrics algorithms that has attracted increasing interest in the chemometrics community in recent years. The goal of MPA is to extract statistical information from the model, towards better understanding of the chemical data. Two key elements of MPA are random sampling and statistical analysis. The core idea of MPA is quite universal with potential applications in the fields, such as chemoinformatics, biostatistics and bioinformatics. In this article, we review the development of MPA in chemometrics. We first present the key elements of MPA. Then, the application of MPA in chemometrics is discussed, such as variable selection, model evaluation, outlier detection, applicability domain definition and so on. Finally, the potential application areas of MPA in future research are prospected. •Model population analysis (MPA) is a framework for chemometrics algorithms.•We present the key elements of MPA.•We discuss the application of MPA in chemometrics.•We prospect the potential application areas of MPA.