Is soft independent modeling of class analogies a reasonable choice for supervised pattern recognition?

Rácz, Anita; Gere, Attila; Bajusz, Dávid; Héberger, Károly

doi:10.1039/c7ra08901e

Cited by 36 publications

(18 citation statements)

References 30 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…The fact that LDA was developed by statisticians, whereas SIMCA was developed by chemists (chemometricians) might contribute to the characteristic differences between their theoretical backgrounds. For example, SIMCA does not require any distributional assumptions, whereas LDA assumes normal distribution and equal variances for each class [37]. In addition, Nikita et al observed that QDA does not give better results than LDA and does not offer an alternative to LDA [38].…”

Section: Chemometrics For Olive Oil Classification According To Its Qmentioning

confidence: 99%

Headspace Gas Chromatography Coupled to Mass Spectrometry and Ion Mobility Spectrometry: Classification of Virgin Olive Oils as a Study Case

García-Nicolás

Arroyo‐Manzanares

Arce

et al. 2020

Foods

View full text Add to dashboard Cite

Due to its multiple advantages, ion mobility spectrometry (IMS) is being considered as a complementary technique to mass spectrometry (MS). The goal of this work is to investigate and compare the capacity of IMS and MS in the classification of olive oil according to its quality. For this purpose, two analytical methods based on headspace gas chromatography (HS-GC) coupled with MS or with IMS have been optimized and characterized for the determination of volatile organic compounds from olive oil samples. Both detectors were compared in terms of sensitivity and selectivity, demonstrating that complementary data were obtained and both detectors have proven to be complementary. MS and IMS showed similar selectivity (10 out of 38 compounds were detected by HS-GC-IMS, whereas twelve compounds were detected by HS-GC-MS). However, IMS presented slightly better sensitivity (Limits of quantification (LOQ) ranged between 0.08 and 0.8 µg g−1 for HS-GC-IMS, and between 0.2 and 2.1 µg g−1 for HS-GC-MS). Finally, the potential of both detectors coupled with HS-GC for classification of olive oil samples depending on its quality was investigated. In this case, similar results were obtained when using both HS-GC-MS and HS-GC-IMS equipment (85.71 % of samples of the external validation set were classified correctly (validation rate)) and, although both techniques were shown to be complementary, data fusion did not improve validation results (80.95% validation rate).

show abstract

Section: Chemometrics For Olive Oil Classification According To Its Qmentioning

confidence: 99%

Headspace Gas Chromatography Coupled to Mass Spectrometry and Ion Mobility Spectrometry: Classification of Virgin Olive Oils as a Study Case

García-Nicolás

Arroyo‐Manzanares

Arce

et al. 2020

Foods

View full text Add to dashboard Cite

show abstract

“…In SIMCA, it is important to select the optimal number of PC for each class (fresh and expired class). The optimal number of PCs can be chosen based on explained variance or determined by (double) crossvalidation [23]. In recent work, Basri et al utilized two parameters namely predicted residual error sum of squares (PRESS) and explained variance to select optimal number of PCs [24].…”

Section: Development Of Simca Classification Modelmentioning

confidence: 99%

Identification of fresh and expired ground roasted robusta coffee using UV-visible spectroscopy and chemometrics

Yulia

Suhandy

2018

MATEC Web Conf.

View full text Add to dashboard Cite

The freshness of ground roasted coffee escapes extremely fast. For this reason, the evaluation of conservation state of ground roasted coffee must be taken into account for acceptability of coffee. Unfortunately, it is difficult to discriminate the fresh and expired ground roasted coffee physically by our naked eyes. Thus, it is desired to develop an analytical method to evaluate the fresh and expired ground roasted coffee using reliable methods. The objective of this research was to evaluate the potential of UVvisible spectroscopy and chemometrics method for classification of fresh and expired ground roasted robusta coffee. A number of 200 samples of robusta fresh coffee and 200 samples of robusta expired coffee was used. The spectral data were pre-treated using standard normal variate (SNV), moving average smoothing (window: 9) and Savitzky-Golay 2 nd derivative (order: 2; window: 11). The analysis data was done statistically using multivariate chemometric techniques, including principal component analysis (PCA) and soft independent modeling of class analogy (SIMCA) in the spectral range of 230-400 nm. PCA with PC1 = 94% and PC2 = 4% showed clear clustering of samples (p ⩽ 0.05). UV-visible spectroscopy with SIMCA analysis allowed to classify between fresh and expired ground roasted robusta coffee with a correct classification rate of 100%.

show abstract

“…This is because the method groups objects together based on applying a PCA to each class of the training set. The ideal number of PCs can be determined by either double cross-validation or amount of explained variance or in some cases, it may be predetermined (Rácz et al,2018). Although SIMCA is a class-modelling technique, it is commonly used as a discriminatory tool in chemometrics.…”

Section: Qualitative Methodsmentioning

confidence: 99%

“…Although SIMCA is a class-modelling technique, it is commonly used as a discriminatory tool in chemometrics. This is warned against by a meta-analysis conducted by Rácz et al (2018), which shows that SIMCA was repeatedly outperformed for the task of discrimination by 29 different methods which includes the majority of the major categories of the available classification methods, such as linear and quadratic discriminant analysis (LDA), Classification and Regression Tree analysis (CART), PLS-DA, k-NN, to name a few.…”

Section: Qualitative Methodsmentioning

confidence: 99%

Unravelling the complexities of wine: A big data approach to yeast assimilable nitrogen

2019

View full text Add to dashboard Cite

Aim: Assimilable Nitrogen (YAN) has been identified as one of the main drivers of wine quality, influencing the production of various aromas and ensuring a successful fermentation to dryness. Due to the number of factors affecting YAN concentration and composition, paired with the complexities of yeast metabolism, more data is required to enable a comprehensive understanding of this important component of the grape juice matrix. The use of high throughput and information-rich techniques such as InfraRed spectroscopy can lead to a fast generation of a large amount of data. In addition, there is a possibility to maximise the information output of the generated data when combined with various descriptive and exploratory statistical techniques.Conclusion: Given the recent developments in the fields of analytical equipment and chemometrics, the review explores the possibility of a Big Data approach for the research of one of the most important and versatile grape juice parameters, namely YAN

show abstract

Is soft independent modeling of class analogies a reasonable choice for supervised pattern recognition?

Abstract: A thorough survey of classification data sets and a rigorous comparison of classification methods show the unambiguous superiority of other techniques over soft independent modeling of class analogies (SIMCA – one class modeling) for classification.

Cited by 36 publications

References 30 publications

Headspace Gas Chromatography Coupled to Mass Spectrometry and Ion Mobility Spectrometry: Classification of Virgin Olive Oils as a Study Case

Headspace Gas Chromatography Coupled to Mass Spectrometry and Ion Mobility Spectrometry: Classification of Virgin Olive Oils as a Study Case

Identification of fresh and expired ground roasted robusta coffee using UV-visible spectroscopy and chemometrics

Unravelling the complexities of wine: A big data approach to yeast assimilable nitrogen

Contact Info

Product

Resources

About