Low-field nuclear magnetic resonance (LF-NMR) combined with chemometrics was used for the rapidly and nondestructively quantitative determination of carbonyl value (CV) frying oils during process. The quality indices process were analyzed as well changes its LF-NMR signal, followed by exploring correlation between relaxation-characteristic investigating influences modeling data preprocessing methods on performance. Based echo decay curves data, prediction model CV established partial least...

Based on fatty acid data, a new rapid method combining data fusion and chemometrics was established to detect peanut oils (PNO) adulteration qualitatively quantitatively. The datasets from overall sn-2 profiles of 350 PNO samples were constructed by serial, parallel, triglyceride fusions. Linear discrimination analysis (LDA) support vector machine classification (SVC) used identify adulteration. model parameters SVC optimized using the grid search particle swarm optimization algorithm....

Abstract The effects of water content, acid value, and peroxide value on interesterification catalyzed by potassium glycerolate (in situ KOH/glycerol) were investigated using lard as a model fat. SEM analysis KOH/glycerol powder showed that numerous 0.5‐ to 5‐μm porous structures formed may play an important role in the reaction. Water content (up 10 %, oil weight) (4.29 7.11 mmol/kg) significantly extended induction period but complete randomization was still achievable. However, when...

Milk is one of the preferred beverages in modern healthy diets, and its freshness great significance for product sales applications. By combining two-dimensional (2D) correlation spectroscopy technique chemometrics, a new method based on visible/near-infrared (Vis/NIR) was proposed to discriminate milk. To clarify relationship be-tween milk spectra, changes physicochemical indicators during storage were analyzed as well Vis/NIR spectra 2D-Vis/NIR spectra. The threshold-value method, linear...

A new hybrid algorithm named EODT-LS-SVR based on least squares support vector regression (LS-SVR) with wavelet-based EODT as preprocessed tools is proposed for removing the interferences and developing quantitative models high precision in near-infrared (NIR) spectra. composed of two steps. In first step, preprocessing EODT, which combines ideas wavelet packet transform (WPT), orthogonal signal correction (OSC) information theory, employed characteristic extraction analyte through...

Synergistic therapy that combines drug/siRNA for the solid tumor treatment is becoming more and popular to reduce side effects enhance antitumor efficiency. Given that, a novel injectable supramolecular hydrogel mPECT(D)/GDDC-4(R)/α-CDgel (mPECT(D) represents DOX-loaded methoxy poly(ethylene glycol)-b-poly(ε-caprolactone-co-1,4,8-trioxa[4.6]spiro-9-un-decanone) (mPECT) nanoparticles, GDDC-4(R) siBcl-2-loaded PG-P(DPAx-co-DMAEMAy)-PCB (GDDC-4) complexes nanoparticles (NPs)) was prepared via...

A new hybrid algorithm (WPNOSC-NPLS) is proposed to eliminate the interference of temperature variation and develop robust calibration models for determining main components concentration in milk using near infrared spectra. At first, three dimensional (3D) spectra including wavelength, absorbency are constructed. Through unfolding method, wavelet packet transform employed split into different frequency components, which can be refold corresponding 3D sub-matrices. Then, orthogonal signal...

Abstract Natural standard grade γ‐tocopherol is of great interest due to its good antioxidant activity and special pharmacological functions. This study separated purified from mixed tocopherols using eco‐friendly low toxicity solvents. γ‐Tocopherol (0.28 g) was obtained per performance with a purity (98.89 ± 0.68)% recovery rate (93.23 0.89)%. The purification conditions were as follows: elution solvent n ‐hexane/ethyl acetate 94.5:5.5 (v/v), sample load 0.5 g, column height diameter ratio...

To take advantages of multiscale property near infrared (NIR) spectra, a new hybrid algorithm (GA-WPLS) was proposed for developing the multivariate regression model in wavelet domain instead spectra domain. At first, packet transform (WPT) and its reconstruction are employed to split raw into different frequency components Then prediction models developed by WPT-based partial least squares (WPLS) where each component is characterized weighted coefficient. Through performance comparison...

A new hybrid algorithm (EWPCS) was proposed for selecting appropriate wavelet packet components containing the variations of analyte as input data regression model based on transform (WPT) and information entropy theory. At first, WPT its reconstruction are employed to split raw spectra into different frequency with maximum levels. Then differences between each component calculated, showing importance component. last, an optimized threshold value determined by performance model, representing...

To improve the training efficiency of least squares-support vector machine (LS-SVM) method, a new algorithm was proposed for developing multivariate regression model using near-infrared (NIR) spectra and named as PCA-PSO-LS-SVM. Coupled with principal component analysis (PCA) particle swarm optimization (PSO), this can take advantage spectral dimension reduction parameter LS-SVM model. In PCA-PSO-LS-SVM algorithm, PCA is firstly employed to reduce raw simplicity based on cumulative...

To efficiently make use of the temperature information in near-infrared (NIR) spectra, a new hybrid algorithm named as WP-WNPLS is proposed to improve prediction ability partial least square (PLS) based regression model.In WP-WNPLS, discrete wavelet packet transform (DWPT) was firstly applied decompose 3-D NIR spectra into series frequency components.In each component, sub-model obtained through using N-way PLS (NPLS) regression.Then, weighted strategy employed take advantage multi-scale...

A new algorithm (WFCE) was proposed for simultaneously eliminating background and noise based on wavelet packet transform (WPT) information entropy theory. At first, WPT reconstruction were employed to split the raw spectra into different frequency components. Then of each component calculated, showing uncertainty measured analyte concentration. last, comparison entropy, importance whole evaluated suitable components representing can be determined removal. WFCE validated by measuring...

A new hybrid algorithm (MIE-WPTOSC), which is the combination of wavelet packet transform (WPT), orthogonal signal correction (OSC) and maximum information extraction (MIE), proposed for interferences elimination in near-infrared (NIR) spectra. In MIE-WPTOSC algorithm, WPT firstly employed de-noising by threshold method, then MIE applied to remove baseline spectra based on theory. At last, uncorrelated concentrations analyte eliminated OSC each frequency band To validate effectiveness a real...

In order to efficiently extract the analyte information in near infrared (NIR) spectra, a new algorithm, which is named as GA-FD-SEL, was proposed by signal processing frequency domain. At first, take advantages of multiscale property NIR wavelet packet transform (WPT) algorithm adopted split raw spectra into series components The optimum decomposition level determined based on entropy contribution each components. Then set WPT initially selected according again, and further using genetic...