Support vector machines (SVM) in near infrared (NIR) spectroscopy: Focus on parameters optimization and model interpretation

Devos, O.; Ruckebusch, Cyril; Durand, Alexandra; Duponchel, Ludovic; Huvenne, J.P.

doi:10.1016/j.chemolab.2008.11.005

Cited by 233 publications

(129 citation statements)

References 13 publications

Supporting

Mentioning

125

Contrasting

Order By: Relevance

“…Support vector machines (SVMs) were used for classification modeling in the study. SVMs are proven to present good generalization performance and to be able to model complex nonlinear boundaries through the use of adapted kernel functions [38]. In this work, radial basis function was used and particle swarm optimization (PSO) based on fivefold cross-validation was adopted for optimization of the parameters in SVM [39][40][41].…”

Section: Rois Extraction Based On Pca and Maskingmentioning

confidence: 99%

Early Detection of Aspergillus parasiticus Infection in Maize Kernels Using Near-Infrared Hyperspectral Imaging and Multivariate Data Analysis

et al. 2017

View full text Add to dashboard Cite

Fungi infection in maize kernels is a major concern worldwide due to its toxic metabolites such as mycotoxins, thus it is necessary to develop appropriate techniques for early detection of fungi infection in maize kernels. Thirty-six sterilised maize kernels were inoculated each day with Aspergillus parasiticus from one to seven days, and then seven groups (were determined based on the incubated time. Another 36 sterilised kernels without inoculation with fungi were taken as control (DC). Hyperspectral images of all kernels were acquired within spectral range of 921-2529 nm. Background, labels and bad pixels were removed using principal component analysis (PCA) and masking. Separability computation for discrimination of fungal contamination levels indicated that the model based on the data of the germ region of individual kernels performed more effectively than on that of the whole kernels. Moreover, samples with a two-day interval were separable. Thus, four groups, DC, D 1-2 (the group consisted of D 1 and D 2 ), D 3-4 (D 3 and D 4 ), and D 5-7 (D 5 , D 6 , and D 7 ), were defined for subsequent classification. Two separate sample sets were prepared to verify the influence on a classification model caused by germ orientation, that is, germ up and the mixture of germ up and down with 1:1. Two smooth preprocessing methods (Savitzky-Golay smoothing, moving average smoothing) and three scatter-correction methods (normalization, standard normal variate, and multiple scatter correction) were compared, according to the performance of the classification model built by support vector machines (SVM). The best model for kernels with germ up showed the promising results with accuracies of 97.92% and 91.67% for calibration and validation data set, respectively, while accuracies of the best model for samples of the mixed kernels were 95.83% and 84.38%. Moreover, five wavelengths (1145, 1408, 1935, 2103, and 2383 nm) were selected as the key wavelengths in the discrimination of fungal contamination levels. In general, near-infrared hyperspectral imaging can be used for early detection of fungal contamination in maize kernels.

show abstract

Section: Rois Extraction Based On Pca and Maskingmentioning

confidence: 99%

Early Detection of Aspergillus parasiticus Infection in Maize Kernels Using Near-Infrared Hyperspectral Imaging and Multivariate Data Analysis

et al. 2017

View full text Add to dashboard Cite

show abstract

“…In addition, specifying parameters G and C is the key step in the SVM because their combined values determine the boundary complexity and thus the classification performance. 36 In our research, the kernel width G was set to 0.15 and the penalty parameter was set to 100. 30 For the implementation of the training and modeling procedure, we employed the existing SVM library LIBSVM presented by Chang and Lin.…”

Section: Support Vector Machinementioning

confidence: 99%

Long-term effects of land use/land cover change on surface runoff in urban areas of Beijing, China

Sun

et al. 2013

J. Appl. Remote Sens

View full text Add to dashboard Cite

Abstract. The objective of this paper is to present a case study to derive land use/land cover (LULC) maps and investigate the long-term effects of LULC change on surface runoff in the fast urbanizing Beijing city. The LULC maps were derived from Landsat TM/ETM+ imagery (acquired in 1992, 1999, 2006, and 2009) using support vector machine method. A long-term hydrologic impact assessment model was applied to assess the impact of LULC change on surface runoff. Results indicated that the selected study area experienced rapid urbanization from 1992 to 2009. Because of urbanization, from 1992 to 2009, modeled runoff increased 30% for the whole area and 35% for the urban portion. Our results also indicated that the runoff increase was highly correlated with urban expansion. A strong relationship (R 2 ¼ 0.849) was observed between the impervious surface percent and the modeled runoff depth in the study area. In addition, a strong positive relationship was observed between runoff increase and percentage of urban areas (R 2 ¼ 0.997 for the whole area and R 2 ¼ 0.930 for the urban portion). This research can provide a simple method for policy makers to assess potential hydrological impacts of future urban planning and development activities. © The Authors. Published by SPIE under a Creative Commons Attribution 3.0 Unported License. Distribution or reproduction of this work in whole or in part requires full attribution of the original publication, including its DOI.

show abstract

“…The fact that these two methods showed poorer classification performance can be explained by the complexity of the problem where the three classes strongly overlap. The results provided for the SVM (91% test CCR) are the ones obtained by the authors on exactly the same data set (for details regarding SVM parameter optimization, we refer to Reference [9]). Even if the HDDA models show better CCR performance on this data set, the purpose here is more to demonstrate the potential of the method for real-life spectroscopy applications than to perform a formal comparison between the methods.…”

Section: Three-class Nir Data Setmentioning

confidence: 99%

Gaussian mixture models for the classification of high‐dimensional vibrational spectroscopy data

Jacques

Bouveyron

Girard

et al. 2010

Journal of Chemometrics

Self Cite

View full text Add to dashboard Cite

Special Issue: Herman Wold Medal Winners 2007-2009International audienceIn this work, a family of generative Gaussian models designed for the supervised classification of high-dimensional data is presented as well as the associated classification method called High Dimensional Discriminant Analysis (HDDA). The advantages of these Gaussian models are: i) the representation of the input density model is smooth; ii) the data of each class are modeled in a specific subspace of low dimensionality; iii) each class may have its own covariance structure; iv) regularization is coupled to the classification criterion to avoid data over-fitting. To illustrate the abilities of the method, HDDA is applied on complex high-dimensional multi-class classification problems in mid-infrared and near infrared spectroscopy and compared to state-of-the-art methods

show abstract

Support vector machines (SVM) in near infrared (NIR) spectroscopy: Focus on parameters optimization and model interpretation

Cited by 233 publications

References 13 publications

Early Detection of Aspergillus parasiticus Infection in Maize Kernels Using Near-Infrared Hyperspectral Imaging and Multivariate Data Analysis

Early Detection of Aspergillus parasiticus Infection in Maize Kernels Using Near-Infrared Hyperspectral Imaging and Multivariate Data Analysis

Long-term effects of land use/land cover change on surface runoff in urban areas of Beijing, China

Gaussian mixture models for the classification of high‐dimensional vibrational spectroscopy data

Contact Info

Product

Resources

About