Principal Component Regression Algorithms Research Articles

A study of entrepreneurial management performance assessment method based on multivariate statistical analysis

Abstract This paper uses multivariate statistical analysis to explore the ordinary least squares regression algorithm, principal component regression algorithm, and partial least squares regression algorithm. Then, an index system for entrepreneurial management performance assessment was constructed with public innovation space, resource collocation and entrepreneurial performance indicators, and an entrepreneurial management performance mediation model was established with this. Finally, the quartiles of entrepreneurial performance's principal components were derived by studying the relationship between entrepreneurial performance assessment and each input. The results show that the first principal component Z 1 of entrepreneurial performance is significant at all quartiles, while the second principal component Z 2 is significant only at the 0.6 quartiles, and the constant terms are significant at the 0.6, 0.7, and 0.8 quartiles. Multivariate statistical analysis is employed in this paper to provide a reference for entrepreneurship management and to assess entrepreneurial performance more accurately.

A Distributed Adaptive Monitoring Method for Performance Indicator in Large-Scale Dynamic Process

The dynamic time-varying characteristic has brought great challenges to the plant-wide process monitoring. In this paper, a distributed adaptive principal component regression algorithm is proposed for the online indicator monitoring of large-scale dynamic process. Firstly, the distributed data subblocks are constructed according to the process operation units. In each subblock, an adaptive re-sampling method based on the subblock data and plant-wide data is presented to construct the modeling sample sets, which can extract the process local and global information simultaneously. Afterwards, the indicator-related feature is extracted, and the Bayesian method is used to integrate the subblock monitoring results. Through the collaborative monitoring of the process local and global feature spaces, a refined monitoring decision can be obtained. Finally, a numerical example and Tennessee Eastman process are used to illustrate the effectiveness of the proposed method.

Application of Data Visualization and Big Data Analysis in Intelligent Agriculture

Intelligent agriculture can renovate agricultural production and management, making agricultural production truly scientific and efficient. The existing data mining technology for agricultural information is powerful and professional. But the technology is not well adapted for intelligent agriculture. Therefore, this paper introduces data visualization and big data analysis into the application scenarios of intelligent agriculture. Firstly, an intelligent agriculture data visualization system was established, and the RadViz data visualization method was detailed for intelligent agriculture. Moreover, the intelligent agriculture data were processed using dimensionality reduction through principal component analysis (PCA) and further optimized through k-means clustering (KMC). Finally, the crop yield was predicted using the multiple regression algorithm and the residual principal component regression algorithm. The crop yield prediction model was proved effective through experiments.

Source apportionment of environmental combustion sources using excitation emission matrix fluorescence spectroscopy and machine learning

The link between particulate matter (PM) air pollution and negative health effects is well-established. Air pollution was estimated to cause 4.9 million deaths in 2017 and PM was responsible for 94% of these deaths. In order to inform effective mitigation strategies in the future, further study of PM and its health effects is important. Here, we present a method for identifying sources of combustion generated PM using excitation-emission matrix (EEM) fluorescence spectroscopy and machine learning (ML) algorithms. PM samples were collected during a health effects exposure assessment panel study in Seattle. We use archived field samples from the exposure study and the associated positive matrix factorization (PMF) source apportionment based on X-ray fluorescence and light absorbing carbon measurements to train convolutional neural network and principal component regression algorithms. We show EEM spectra from cyclohexane extracts of the archived filter samples can be used to accurately apportion mobile and vegetative burning sources but were unable to detect crustal dust, Cl-rich, secondary sulfate and fuel oil sources. The use of this EEM-ML approach may be used to conduct PM exposure studies that include source apportionment of combustion sources.

Modeling ordered subcarrier SNR in MIMO-OFDM wireless links

Recent research has shown that the detected bit error-rate of coded and bit-interleaved multiple-input, multiple-output orthogonal frequency division multiplexing (MIMO-OFDM) links is characterized by the ordered post-processing signal-to-noise ratio (SNR). In this paper, we show that the per-stream ordered SNR converges asymptotically (with delay spread) to the inverse, marginal cumulative distribution function (CDF) of post-processing SNR. We derive a new approximation of the post-processing SNR inverse marginal CDF through an approximation of the marginal CDF for ordered eigenvalues in Wishart random matrices. By evaluating the independence of subcarriers in terms of the power delay profile, we show that the inverse marginal CDF is also an important characterization of post-processing SNR in practice. We exploit this structure to improve principal component regression algorithms that empirically infer a low-dimension basis for ordered SNR. Numerical simulations show high accuracy of this basis with just three dimensions for arbitrary power delay profiles. The consequences of this work include high-resolution limited channel feedback, simpler channel models for system simulation and algorithm design, as well as the reduction of dimension in link quality metrics for link adaptation. Simulations demonstrate the application to link adaptation in IEEE 802.11n with constellations that may be adapted per spatial stream.

New HPLC‐chemometric approaches to the analysis of isoflavones in Trifolium lucanicum Gasp.

New HPLC-chemometric approaches were proposed for the simultaneous chromatographic quantification of daidzein, genistein, formononetin, and biochanin A in the samples consisting of the aerial parts of Trifolium lucanicum Gasp. (Leguminosae). Partial least squares and principal component regression algorithms were applied to the multiple chromatographic data set obtained by measuring at 240, 248, 256, and 264 nm to construct HPLC-partial least squares and HPLC-principal component regression calibrations. Chromatographic separation was carried out by using a mobile phase containing methanol, acetate buffer (pH=4.75) and acetonitrile (21:58:21, v/v/v) on the reversed phase column, Supelcosil™ LC-18 (15 cm×4.6 mm id). In addition, conventional HPLC based on the detection at a single wavelength was used for the determination of each compound in the extracts of T. lucanicum. The validity and applicability of the proposed HPLC-chemometric and conventional HPLC methods were performed by analyzing various synthetic plant samples. A good agreement was observed in the application of the proposed HPLC-chemometric tools to the synthetic and extracted samples of T. lucanicum.

Near Infrared Reflectance Spectroscopy as a Tool for the Determination of Dichloromethane Extractable Matter and Moisture Content in Combed Wool Slivers

In the wool textile field the quantitative determination of solvent extractable matter and moisture content is a crucial analysis for the evaluation of combed sliver quality. The test carried out includes the acquisition of a series of data about dichloromethane soluble matter (according to the International Wool Textile Organisation—IWTO 10–01 specification), and moisture content in combed wool slivers and the search for a correlation between these data and the near infrared spectra of samples. Combed wool slivers tested were of different origins and variable mean diameters and were obtained from different combing mills in the industrial district of Biella, the principal wool textile region of Italy. The spectrophotometer used was a FT-NIR system (Perkin-Elmer Spectrum IdentiCheck). Spectra were collected in the region from 3700 to 10,000 cm−1 in reflection mode. The extraction of greasy matter from wool tops was carried out with a continuous extraction technique on a 10 g wool sample with a total extraction time of about 3 hours in a soxhlet apparatus. The results express the weight (obtained by the mean of two determinations) of the dichloromethane soluble extract as a percentage of the dry weight of the de-greased sample. For the determination of dichloromethane extractable matter, 103 samples were used for calibration. Some wool samples were deliberately under-scoured and others were re-scoured in a combing mill in order to obtain a wide range of data, ranging from 1.15% to 0.21%. Spectra were analysed using Quant+ (Perkin-Elmer Software). The best results were obtained with the PLS1 (Partial Least Square) algorithm when considering the spectral region 9000–3800 cm−1 [Standard Error of Prediction ( SEP) = 0.1042, mean value ( M) = 0.6963%, Coefficient of Determination ( R2) = 0.85]. A cross-validation was used. The determination of moisture content in combed wool sliver was performed by drying wool (80 g for a single determination) in an oven to constant weight using forced air at 105 ± 2°C. 85 samples were used for the calibration. Deliberate variations in regain were induced by exposing the wool to a dry or moist atmosphere. In this way moisture contents ranging from 9% to 15% were obtained. NIR spectra were analysed using Quant+ Software. A cross-validation was used. The best results were obtained with the Principal Component Regression algorithm when considering the spectral region 9000–3800 cm−1. A SEP of 0.4954 ( M = 11.76%) and a R2 of 0.90 were found. A limited number of determinations of grease and moisture content were carried out using the models obtained and compared with values determined manually.

SIMULTANEOUS RESOLUTION OF DEXAMETHASONE AND POLYMYXIN B BY SPECTROPHOTOMETRY DERIVATIVE AND MULTIVARIATE METHODS

Five spectrophotometric methods are described and applied to resolve binary mixtures of the corticosteroid Dexamethasone and the antibiotic Polymyxin B. The simultaneous determination of the two compounds was firstly accomplished by two derivative methods using the “first derivative” and the “ratio spectrum-zero crossing derivative”, which were satisfactory used to determine synthetic mixtures of these compounds in different ratios. Absorption spectra of compounds were used to optimise the spectral data set by performing the calibration by partial least squares-1 and 2 and principal component regression algorithms. These calibration models were evaluated by internal validation (prediction of compounds in its own designed training set of calibration), by cross-validation (obtaining statistical parameters that show the efficiency for a calibration fit model) and by external validation over synthetic mixtures. The five described procedures do not require any separation step. Repeatability and reproducibility studies were achieved over two series of ten standards for Dexamethasone and Polymyxin B showing no significant differences at 95% confidence level in the four spectrophotometric methods. A comparison of the derivative and multivariate calibration results obtained on synthetic mixtures was performed resulting in agreement of the values obtained.

Spectrophotometric resolution of ternary mixtures of Tartrazine, Patent Blue V and Indigo Carmine in commercial products

Four spectrophotometric methods are described and applied to resolve ternary mixtures of the food dyes Tartrazine (E-102), Patent Blue V (E-131) and Indigo Carmine (E-132). The simultaneous determination of these three dyes was firstly accomplished by two derivative methods using the “first derivative” and the “ratio spectrum-zero crossing derivative”, which were satisfactorily used to determine synthetic mixtures of these colorants in different ratios and in three commercial products. Direct absorption spectra of compounds were used to optimize the spectral data set to perform the calibration by partial least squares-1 and principal component regression algorithms. Both calibration models were evaluated by internal validation (prediction of dyes concentration in its own designed training set of calibration), by cross-validation (obtaining statistical parameters that show the efficiency for a calibration fit model) and by external validation over synthetic and commercial mixtures. The four described procedures do not require any separation step. Repeatability and reproducibility studies were achieved over two series of nine standards for each dye showing no significant differences at 95% confidence level in the four spectrophotometric methods. A comparison of the two derivative and the two multivariate calibration results obtained in commercial products was performed resulting in a good agreement of the values obtained.

Evanescent Fiber-Optic Chemical Sensor for Monitoring Volatile Organic Compounds in Water

The transport of trichloroethylene, 1,1,1-trichloroethane, and toluene in aqueous solutions through a polydimethylsiloxane film was modeled using a Fickian diffusion model to fit data obtained from an evanescent fiber-optic chemical sensor (EFOCS). The resultant diffusion coefficients for these analytes were respectively 3 × 10(-)(7), 5 × 10(-)(7), and 1 × 10(-)(7) cm(2)/s. Inclusion of an interfacial conductance term, defined as the ratio of the mass transport coefficient across the polymer surface and the analyte diffusion coefficient in the polymer, was required to accurately model the data. It was determined that the interfacial conductance terms were generally of the same order of magnitude for the analytes examined, suggesting a constant transport mechanism for the analytes. Linear chemometric algorithms were used to model the EFOCS response to aqueous mixtures of the three analytes with individual analyte concentrations between 20 and 300 ppm. Both partial least-squares and principal component regression algorithms performed comparably on the calibration sets, with cross-validated root-mean-squared errors of prediction for trichloroethylene, 1,1,1-trichloroethane, and toluene of approximately 26, 29, and 22 ppm, respectively. The resultant prediction model was then used to determine analyte concentrations in an independent data set with comparable precision.