The successive projections algorithm for variable selection in spectroscopic multicomponent analysis

Araújo, Mário César Ugulino de; Saldanha, Teresa Cristina Bezerra; Galvão, Roberto Kawakami Harrop; Yoneyama, Takashi; Chame, Henrique C.; Visani, Valeria

doi:10.1016/s0169-7439(01)00119-8

Cited by 1,093 publications

(595 citation statements)

References 23 publications

Supporting

Mentioning

539

Contrasting

Unclassified

Order By: Relevance

“…The multi-spectral device can only contain few spectral bands, so feature selection us been used, which means locating the best minimum subset of the original features. This was done using the Successive Projection Algorithm (SPA) (Araujo et al, 2001), first, the instrumental response data are used to create chains of variables according to a sequence of vector projection operations designed to minimize multi-collinearity among the variables of the chain. Second, the algorithm creates a model for each of the candidate subsets of variables extracted from the chains generated.…”

Section: Spectral Data Analysis For Disease Detectionmentioning

confidence: 99%

Multispectral band selection for imaging sensor design for vineyard disease detection: case of Flavescence Dorée

Al-Saddik

Simon

Brousse³

et al. 2017

Advances in Animal Biosciences

View full text Add to dashboard Cite

Disease detection and control is thus one of the main objectives of vineyard research in France. Monitoring diseases manually is fastidious and time consuming, so current research aims to develop an automatic detection of vineyard diseases. This project explored the use of a high-resolution multi-spectral camera embedded on a UAV (Unmanned Aerial Vehicle) to identify the infected zones in a field. In-field spectrometry studies were performed to identify the best spectral bands for the sensor design. The best models were found to be the function of the grapevine variety considered and the 520-600-650-690-730-750-800 nm bands were found to be the most efficient for all types of grapevines, with an overall classification accuracy of more than 94%.

show abstract

Section: Spectral Data Analysis For Disease Detectionmentioning

confidence: 99%

Multispectral band selection for imaging sensor design for vineyard disease detection: case of Flavescence Dorée

Al-Saddik

Simon

Brousse³

et al. 2017

Advances in Animal Biosciences

View full text Add to dashboard Cite

show abstract

“…Each chain starts with one of the variables under consideration and is successively augmented with additional variables chosen in order to display the least collinearity with the previous ones, as described in earlier papers. 14,17 The notation {SEL(1, k), SEL(2, k), …, SEL(M, k)} is used to denote the index set of variables belonging to the chain initialized with x k (that is, SEL(1, k) = k).…”

Section: Robust Variable Selection By Spamentioning

confidence: 99%

“…13 This algorithm was originally proposed for the minimization of collinearity problems in Multiple Linear Regression (MLR). 14 Honorato et al 13 adapted SPA to select variables that convey information concerning the property of interest and are robust with respect to the differences between two instruments. The proposed strategy was applied to two calibration transfer problems involving the determination of T90% in gasoline by FT-IR spectrometry and moisture in corn by Near Infrared (NIR) spectrometry.…”

Section: Introductionmentioning

confidence: 99%

Multivariate calibration transfer employing variable selection and subagging

Martins¹,

Galvão²,

Pimentel³

2010

J. Braz. Chem. Soc.

View full text Add to dashboard Cite

Neste artigo, é proposta uma nova técnica para transferência de calibração, que combina o Algoritmo das Projeções Sucessivas (APS) para seleção de variáveis robustas com a técnica de sub-amostragem e agregação de modelos conhecida como subagging. A técnica proposta tem por objetivo construir modelos de Regressão Linear Múltipla (RLM) que sejam robustos com respeito a diferenças na resposta instrumental de dois espectrômetros (primário e secundário). Para isso, um pequeno conjunto de amostras de transferência, com espectros adquiridos no instrumento secundário, é empregado para guiar o procedimento de seleção de variáveis. A eficiência da técnica proposta é demonstrada em um estudo de caso envolvendo a determinação por FT-IR da massa específica e duas temperaturas de destilação (T10%, T90%) em amostras de gasolina e a determinação por NIR de umidade em amostras de milho. Em termos do erro quadrático médio de predição no espectrômetro secundário, os modelos RLM gerados pela abordagem APS-subagging forneceram resultados melhores que os obtidos por Mínimos Quadrados Parciais empregando Padronização Direta por Partes. Em particular, o uso de subagging resultou em uma redução mais sistemática do erro de predição com a inclusão progressiva de amostras de transferência. This paper proposes a new technique for calibration transfer, which combines the Successive Projections Algorithm (SPA) for robust variable selection with the subsampling and model aggregation technique known as subagging. The proposed technique is aimed at building Multiple Linear Regression (MLR) models that are robust with respect to differences in the instrumental response of two spectrometers (primary and secondary). For this purpose, a small set of transfer samples with spectra acquired at the secondary instrument is employed to guide the variable selection procedure. The efficiency of the proposed technique is demonstrated in a case study concerning the FT-IR determination of specific mass and two distillation temperatures (T10%, T90%) for gasoline samples and the NIR determination of moisture in corn samples. In terms of the root-mean-square error of prediction at the secondary spectrometer, the MLR models obtained according to the SPA-subagging approach provided better results in comparison with Partial Least Squares employing Piecewise Direct Standardization. In particular, the use of subagging resulted in a more systematic reduction in the prediction error with the progressive inclusion of transfer samples.Keywords: multivariate calibration, calibration transfer, variable selection, successive projections algorithm, subagging IntroductionIn a wide sense, the term Calibration Transfer concerns a body of methods aimed at compensating for changes in experimental conditions that would compromise the prediction accuracy and reliability of a multivariate model. Such changes may refer to physical/ chemical characteristics of the sample (such as viscosity, granularity, surface texture, and presence of interferent species), environmental conditions (te...

show abstract

“…More details concerning the operations involved in SPA can be found elsewhere. 1,9 In all previous applications of SPA, the performance metric employed in Phases 2 and 3 was the RMSEV value obtained in an independent validation set of N val samples, defined as (1) where y val,n and y^v al,n are the reference and predicted values of the parameter under consideration for the n th validation sample. In the present work, an extension of this criterion to the cross-validation case is adopted by considering the root mean square error of cross-validation (RMSECV) defined as (2) where y cal,n is the reference value of the parameter under consideration for the n th sample of the calibration set itself, which contains N cal samples.…”

Section: Background and Theorymentioning

confidence: 99%

“…[1][2][3][4][5][6] To address this issue, the present paper presents a comparative study between the use of a separate validation set and leaveone-out cross-validation for the selection of spectral variables by SPA. This investigation is of value to determine whether there are gains, in either parsimony or prediction performance, that may justify the use of crossvalidation in SPA in view of the computational overhead 1581Galvão et al Vol.…”

Section: Introductionmentioning

confidence: 99%

Cross-validation for the selection of spectral variables using the successive projections algorithm

Galvão¹,

Araújo²,

Silva³

et al. 2007

J. Braz. Chem. Soc.

Self Cite

View full text Add to dashboard Cite

Este trabalho compara o uso de um conjunto de validação separado e de validação cruzada amostra-a-amostra para guiar a seleção de variáveis no Algoritmo das Projeções Sucessivas (APS) para calibração multivariada. Análises de diesel e milho por espectrometria NIR são apresentadas. Uma interface gráfica do APS encontra-se disponível em www.ele.ita.br/ kawakami/spa/ This work compares the use of a separate validation set and leave-one-out cross-validation to guide the selection of variables in the Successive Projections Algorithm (SPA) for multivariate calibration. Two case studies involving diesel and corn analysis by NIR spectrometry are presented. A graphical interface for SPA is available at www.ele.ita.br/~kawakami/spa/

show abstract

The successive projections algorithm for variable selection in spectroscopic multicomponent analysis

Cited by 1,093 publications

References 23 publications

Multispectral band selection for imaging sensor design for vineyard disease detection: case of Flavescence Dorée

Multispectral band selection for imaging sensor design for vineyard disease detection: case of Flavescence Dorée

Multivariate calibration transfer employing variable selection and subagging

Cross-validation for the selection of spectral variables using the successive projections algorithm

Contact Info

Product

Resources

About