Improved variable reduction in partial least squares modelling based on predictive-property-ranked variables and adaptation of partial least squares complexity

The calibration performance of partial least squares for one response variable (PLS1) can be improved by elimination of uninformative variables. Many methods are based on so-called predictive variable properties, which are functions of various PLS-model parameters, and which may change during the variable reduction process. In these methods variable reduction is made on the variables ranked in descending order for a given variable property. The methods start with full spectrum modelling. Iteratively, until a specified number of remaining variables is reached, the variable with the smallest property value is eliminated; a new PLS model is calculated, followed by a renewed ranking of the variables. The Stepwise Variable Reduction methods using Predictive-Property-Ranked Variables are denoted as SVR-PPRV. In the existing SVR-PPRV methods the PLS model complexity is kept constant during the variable reduction process. In this study, three new SVR-PPRV methods are proposed, in which a possibility for decreasing the PLS model complexity during the variable reduction process is build in. Therefore we denote our methods as PPRVR-CAM methods (Predictive-Property-Ranked Variable Reduction with Complexity Adapted Models). The selective and predictive abilities of the new methods are investigated and tested, using the absolute PLS regression coefficients as predictive property. They were compared with two modifications of existing SVR-PPRV methods (with constant PLS model complexity) and with two reference methods: uninformative variable elimination followed by either a genetic algorithm for PLS (UVE-GA-PLS) or an interval PLS (UVE-iPLS). The performance of the methods is investigated in conjunction with two data sets from near-infrared sources (NIR) and one simulated set. The selective and predictive performances of the variable reduction methods are compared statistically using the Wilcoxon signed rank test. The three newly developed PPRVR-CAM methods were able to retain significantly smaller numbers of informative variables than the existing SVR-PPRV, UVE-GA-PLS and UVE-iPLS methods without loss of prediction ability. Contrary to UVE-GA-PLS and UVE-iPLS, there is no variability in the number of retained variables in each PPRV(R) method. Renewed variable ranking, after deletion of a variable, followed by remodelling, combined with the possibility to decrease the PLS model complexity, is beneficial. A preferred PPRVR-CAM method is proposed.     

Chromatographic separation techniques and data handling methods for herbal fingerprints: a review

As herbal medicines have an important position in health care systems worldwide, their current assessment and quality control are a major bottleneck. Over the past decade, major steps were taken not only to improve the quality of the herbal products but also to develop analytical methods ensuring their quality. Nowadays, chromatographic fingerprinting is the generally accepted technique for the assessment and quality control of herbal products. This paper briefly considers the evolution of the regulations and guidelines on the quality control of herbal medicines, and reviews the established analytical techniques for herbal fingerprinting with an emphasis on the most recent developments, such as miniaturized techniques, new stationary phases, analysis at high temperatures and multi-dimensional chromatography. Accessory to the new analytical techniques, the chemometric data handling techniques applied are discussed. Chemometrics provide scientists with useful tools in understanding the huge amounts of data generated by the analytical advances and prove to be valuable for quality control, classification and modelling of, and discrimination between herbal fingerprints.     

Integrated band intensities in N2O4 in the infrared range

We have measured the integrated band intensities of the ν₉ and ν₁₁ bands of N₂O₄ which are observed around 1757 and 1261 cm^-1, respectively. By varying temperature and pressure, we have obtained: S_band(ν₉) = 9.60(130), 9.10(24), 8.80(66) and S_band(ν₁₁)= 5.93(64), 5.70(21) and 5.33(46) (in 10^-17 cm/molecule) at 293.15 (60), 277.25 (60) and 261.65 (60) K, respectively.     

Absolute line intensities of CO2 in the 3090-3920 cm region

Absolute line intensities of ¹³C¹⁶O₂ were retrieved from high-resolution Fourier transform spectra recorded in the region 3090-3920 cm⁻¹. The uncertainty of the line intensity determination is estimated to be between 3 and 5% for the strong lines. The global fittings of the observed line intensities within the framework of the effective operators approach have been performed, reaching the experimental accuracy. A comparison of newly measured line intensities with those found in the HITRAN database is presented.     

Visualisation of biopolymer mixtures using confocal scanning laser microscopy (CSLM) and covalent labelling techniques

Confocal scanning laser microscopy (CSLM) has been used to study the behaviour of mixtures of proteins, gelatine, whey proteins and β-lactoglobulin, and polysaccharides, dextran, gellan gum, carrageenan, gum Arabic, and starch. CSLM proved to be a suitable technique to visualise the microstructure of these (phase separated) mixtures in two and three-dimensional images. Contrast through fluorescence is obtained either by covalent labelling (polysaccharides and proteins) or non-covalent labelling (proteins and starch). Double and triple labelling allows the visualisation of individual components in a complex mixture of biopolymers.     

Experimental evidence of mounting grooves and serration patterns on fin lines characteristics

In this paper, the mounting grooves and serration patterns of fin line circuits are investigated experimentally. Our measurements show that the short-circuit ended mounting grooves degrade seriously the insertion loss of fin lines at the low frequencies of the band. The open ended mounting grooves are found better for fin lines applications and measurements. An improvement of 0.7 dB is observed over the whole band, when the serration patterns of the fin line are placed into the open-ended mounting grooves. These results can be used for measuring the losses and for designing transitions or other fin line circuits.     

Limits on neutrino oscillations from a study of νe charged current interactions

We have fully analysed all events with a single electron obtained in an exposure of BEBC to the 350 GeV wide-band neutrino beam at CERN. The data agrees well with expectations based on the calculated ν_e flux and assuming μ?e universality. Using both the rate and the kinematic distributions, improved limits are set on ν_e → ν_X, ν_μ → ν_e, ν_μ → ν_τ mixing.     

A capacity constrained single-facility dynamic lot-size model

A computationally efficient algorithm for a multi-period single commodity production planning problem with capacity constraints is developed. The model differs from earlier well-known studies involving concave cost functions in the introduction of production capacity constraints which need not be equal in every period. The objective is to find an optimal production schedule that minimizes the total production and inventory costs. Backlogging is not allowed. The structure of the optimal solution is characterized and then used in an efficient algorithm.