Subject: Mathematics & Computer Science, Other Keywords: Chemometric data,sparse autoencoder, gaussian process regressor, pareto optimization.
Online: 9 May 2019 (11:31:46 CEST)
We proposed a deep learning based chemometric data analysis technique. We trained L2 regularized sparse autoencoder end-to-end for reducing the size of the feature vector to handle the classic problem of curse of dimensionality in chemometric data analysis. We introduce a novel technique of automatic selection of nodes inside hidden layer of an autoencoder through pareto optimization. Moreover, linear regression, ϵ-SVR , and Gaussian process regressor are applied on the reduced size feature vector for the regression. We evaluated our technique on orange juice and wine dataset and results are compared against state-of-the-art methods. Quantitative results are shown on Normalized Mean Square Error (NMSE) and the results show considerable improvement in the state-of- the-art.
ARTICLE | doi:10.20944/preprints202109.0478.v1
Online: 28 September 2021 (21:36:05 CEST)
Edible mushrooms have been recognized as highly nutritional food for a long time, due to their specific flavor, texture and also for therapeutic effects. This study proposes a new simple approach, based on FT-IR analysis, followed by statistical methods, in order to differentiate three wild mushrooms species from Romanian spontaneous flora, namely Armillaria mellea, Boletus edulis and Cantharellus cibarius. The preliminary data treatment consisted of data set reduction with principal component analysis (PCA), which provided scores for the next methods. Linear discriminant analysis (LDA) manage to 100% classify the three species and the cross validation step of the method returned 97.4% of correctly classified samples. Only one A. mellea sample overlapped on B. edulis group. When kNN was used in the same manner as LDA, the overall percent of correctly classified samples from the training step was 86.21%, while for holdout set the percent raised at 94.74%. The lowered values obtained for the training set was due to one C. cibarius sample, two B. edulis and five A. mellea, which were placed to other species. Anyway, for holdout sample set, only one sample from B. edulis was misclassified. The fuzzy c-means clustering (FCM) analysis successfully classified investigated mushroom samples according to their species, meaning that in every partition the predominant specie had the biggest DOMs, while samples belonging to other specie had lower DOMs.
ARTICLE | doi:10.20944/preprints201906.0268.v1
Subject: Biology, Agricultural Sciences & Agronomy Keywords: Extra virgin olive oil; Hojiblanca; Arbequina, antioxidant properties; polyphenols; chemometric analysis.
Online: 26 June 2019 (13:52:33 CEST)
The health benefits of extra virgin olive oil (EVOO) are related with its chemical composition and the presence of bioactive compounds with antioxidant properties. The aim of this study was evaluate antioxidant compounds (pigments, CoQ10 and phenolic compounds) and antioxidant properties of EVOO from the same region comparing different cultivars (Hojiblanca and Arbequina), harvest year and crop stage. Antioxidant properties of oils were studied before and after a gastrointestinal digestion process, by in vitro assays (DPPH, ABTS and FRAP) and antioxidant markers in Caco-2 cells (reactive oxygen species production). The content of bioactive compounds measured was significantly affected by cultivar and harvest year (except for carotenoids) and by the crop stage (except for coenzyme Q10). Higher amount of coenzyme Q10 was observed in Hojiblanca than in Arbequina EVOO. Total phenol content and antioxidant properties were also different depending on cultivar and harvest year and the in vitro digestion process strongly improved antioxidant marker values. Antioxidant potential in bioaccessible fractions was mainly related with content of coenzyme Q10 and phenolic compounds in EVOO. Chemometric analysis showed that the oils were clearly classified by cultivars, harvest and crop stage, according with chemical composition and antioxidant activity analyzed in the present study
ARTICLE | doi:10.20944/preprints201803.0053.v1
Subject: Chemistry, Food Chemistry Keywords: biogenic amines; chemometric analysis; DLLME, GC-MS; storage conditions; stopper type
Online: 7 March 2018 (13:08:12 CET)
1) Background: A survey of biogenic amines profile in opened wine bottles has been established. Opened bottles of red and white wine were submitted to different temperature as well as different kind of stopper (screw cap, cork stopper) and use of vacuum devices. A total of six wine made from different variety of grapes were obtained from Polish vineyard places in different region of Poland; 2) Results: DLLME-GC-MS procedure for biogenic amines determination was validated and applied for wine samples analysis. The total content of BAs in white wines ranged from 442 µg/L to 929 µg/L, while in red wines ranged from 669 µg/L to 2244 µg/L the set of just opened wine samples. The most abundant biogenic amines in the six analysed wines were histamine and putrescine; 3) Conclusion: Considering the commercial availability of the analysed wines, there was no relationship between the presence of biogenic amines in a given wine and their availability on the market. However, it was observed that the different storage conditions employed in this experiment affect not only the biogenic amines profile, but also the pH. The results were confirmed by chemometric analysis.
Subject: Chemistry, Analytical Chemistry Keywords: Amomum villosum Lour.; Amomum villosum Lour. var. xanthioides T. L.Wu et Senjenis; GC-MS; Chemometric Techniques; volatile oil
Online: 1 April 2019 (08:30:57 CEST)
Fructus Amomi (FA) is usually regarded as the dried ripe fruits of Amomum villosum Lour. (FAL) or Amomum villosum Lour. var. xanthioides T. L.Wu et Senjenis (FALX) However, FAL, which always has a much higher price because of its better quality, is confused with FALX in the market. As volatile oil is the main constituent of FA, a strategy of chromatography-mass spectrometry (GC-MS) and chemometric approches was applied to compare the chemical composition of FAL and FALX. The results showed that the oil yield of FAL was significantly higher than that of FALX. Total ion chromatograph (TIC) showed that cis-nerolidol existed only in FALX. Bornyl acetate and camphor, could be considered as the most important ones in FAL and FALX respectively. Moreover, hierarchical cluster analysis (HCA) and principal component analysis (PCA) successfully distinguished the chemical constitutes of the volatile oils in FAL and FALX. Additionally, bornyl acetate, α-cadinol, linalool, β-myrcene, camphor, d-limonene, terpinolene and endo-borneol were selected as the potential markers for discriminating FAL and FALX by partial least squares discrimination analysis (PLS-DA). In conclusion, this present study first developed a scientific approach to separate FAL and FALX based on volatile oils by GC-MS combined with chemometric techniques.
Subject: Keywords: bioprocess models; model validation; model calibration; Quality by Design; mechanistical and statistical models; hybrid models; chemometric models; Biopharmaceutical engineering; regulatory guidance
Online: 10 May 2021 (09:57:09 CEST)
In bioprocess engineering the Qualtiy by Design (QbD) initiative encourages the use of models to define design spaces. However, clear guides on how models for QbD are validated are still missing. In this review we provide a comprehensive overview about validation methods, mathematical approaches and metrics currently applied in bioprocess modeling. The methods cover analytics for data used for modeling, model training and selection, measures for predictiveness and model uncertainties. We point out general issues in model validation and calibration for different types of models and put this into context of existing health authority recommendations. This review provides the start-point for developing a guidance for model validation approaches. There is no one-fits-all approach but this review shall help to identify the best fitting validation method or combination of methods for the specific task and type of bioprocess models that is developed.