RT info:eu-repo/semantics/article
T1 Simultaneous class-modelling in chemometrics: A generalization of Partial Least Squares class modelling for more than two classes by using error correcting output code matrices
A1 Valencia García, Olga
A1 Ortiz Fernández, Mª Cruz
A1 Ruiz Miguel, Santiago
A1 Sánchez Pastor, Mª Sagrario
A1 Sarabia Peinador, Luis Antonio
K1 Compliant class-models
K1 Error-correcting output code (ECOC)
K1 Partial least squares multi-response regression (PLS2)
K1 Sensitivity
K1 Specificity
K1 Diagonal modified confusion entropy (DMCEN)
K1 Química
K1 Chemistry
K1 Matemáticas
K1 Mathematics
AB The paper presents a new methodology within the framework of the so-called compliant class-models, PLS2-CM,designed with the purpose of improving the performance of class-modelling in a setting with more than twoclasses. The improvement in the class-models is achieved through the use of multi-response PLS models with theclasses encoded via Error-Correcting Output Codes (ECOC), instead of the traditional class indicator variablesused in chemometrics.The proposed PLS2-CM entails a decomposition of a class-modelling problem into a series of binary learners,based on a family of code matrices with different code length, which are evaluated to obtain simultaneouscompliant class-models with the best performance.The methodology develops both a new encoding system, based on multi-criteria optimization to search foroptimal coding matrices, and a new decoding system, based on probability thresholds to assign objects to classmodels. The whole procedure implies that the characteristics of the dataset at hand affect the final selection of thecoding matrix and therefore of built class-models, thus giving rise to a data-driven strategy.The application of PLS2-CM to a variety of cases (controlled data, experimental data and repository datasets)results in an enhanced class-modelling performance by means of the suggested procedure, as measured by theDMCEN (Diagonal Modified Confusion Entropy) index and by sensitivity-specificity matrices. The predictiveability of the compliant class-models has been evaluated.
PB Elsevier
SN 0169-7439
YR 2022
FD 2022-08
LK http://hdl.handle.net/10259/7425
UL http://hdl.handle.net/10259/7425
LA eng
NO This work is part of the project with reference BU052P20 financed by Junta de Castilla y Leon, Conserjería de Educacion with the aid of European Regional Development Funds.
DS Repositorio Institucional de la Universidad de Burgos
RD 06-may-2024