Least-squares Regression Method Research Articles

Influence of polarized reflection on airborne remote sensing of canopy foliar nitrogen content

ABSTRACT Radiation detected by sensors can reveal information regarding canopy structure and leaf surface properties. It has been shown that the scattering coefficient (Wλ ), a ‘pure’ spectrum obtained after correction for the effect of canopy structure on the bidirectional reflectance factor (BRF), can provide information about biochemical composition; e.g. nitrogen content, more objectively. However, specular radiation reflected from leaf surface (without any information from the leaf interior) has a potential influence on Wλ and thus consequently affects the estimation of canopy nitrogen content (CNC). The so-called specular reflection influence has rarely been studied. Polarized reflectance (R p), as part of the specular reflectance, was used to assess this kind of influence. To model the accurate R p within the study area, six bidirectional polarization distribution function (BPDF) models were intercompared, using the existing POLDER/PARASOL R p database, to produce R p for three vegetation types: needleleaf, broadleaf, and mixed forest. Then, correction for polarized reflection and canopy structure was made to yield a more reasonable and accurate Wλ . An efficient interval partial least-squares regression (iPLSR) method for CNC estimation was used to analyse the impact of polarized reflection on both the 400 nm–2,500-nm canopy Wλ and CNC estimation. The results showed that the contribution of polarization was greater in the strong-absorption spectral region, with an influence on Wλ estimation of up to 25% and 28% in the visible and short-wave infrared regions, respectively. Also, improvements of 0.93% in average regarding the relative mean square error of cross-validation (RMSECV) were seen in CNC estimation accuracy. Moreover, sensitivity analyses of BPDF model parameters, and of the fixed total number of intervals in the iPLSR were also conducted. Improvements with an average of 1.19% regarding the RMSECV was consistently seen in the accuracy of CNC regression after polarization correction, no matter how many fixed intervals were. The results also gave guides on the selection of spectral regions used to estimate forest CNC.

Measurement and correlation of the saturated vapor pressures of Ethenyltris(2,2,2-trifluoroethoxy)silane, Dimethoxymethyl(3,3,3-trifluoropropyl)silane and Trimethoxy(3,3,3-trifluoropropyl)silane

The saturated temperature of ethenyltris(2,2,2-trifluoroethoxy)silane (trisF-ES), dimethoxymethyl(3,3,3-trifluoropropyl)silane (F-DiMOMS) and trimethoxy(3,3,3-trifluoropropyl)silane (F-TriMOS) were determined using an inclined ebulliometer from 3.00 to 100.00 kPa. The experimental data were fitted with the Antoine equation and the Clarke−Glew equation by a weighted least-squares regression method, and gave the Antoine parameters (A = 8.8461, B = 1207.11, C = −114.06 K, A = 8.8040, B = 1166.06, C = −97.69 K and A = 8.9712, B = 1279.86, C = −92.77 K) and standard vaporization enthalpy ΔlgHm0(298.15 K) = 55.84 kJ mol−1, ΔlgHm0(298.15 K) = 47.54 kJ mol−1 and ΔlgHm0(298.15 K) = 49.48 kJ mol−1 for trisF-ES, F-DiMOMS and F-TriMOS, respectively. The critical properties of trisF-ES, F-DiMOMS and F-TriMOS, including critical temperature, pressure, and volume, were estimated by the group contribution method introduced by Nannoolal et al. The acentric factors (ω) of trisF-ES, F-DiMOMS and F-TriMOS were estimated from these critical parameters and the reduced saturated vapor pressures calculated using both the Antoine equation and the Clarke−Glew equation.

System Identification and Control for a Tail-Sitter Unmanned Aerial Vehicle in the Cruise Flight

This work presents the implementation of system identification and model predictive control for a tail-sitter unmanned aerial vehicle (UAV) in cruise flight. The mathematical model of longitudinal and lateral directions of the UAV has been derived in the state-space form for grey-box modeling. The least-square regression method is augmented with regulation and solved by applying the trust-region algorithm. Outdoor flight tests were conducted to acquire the data for system identification assisted by a signal generator module. The UAV dynamic was sufficiently excited in both longitudinal and lateral directions during the flight test. The flight data were applied to the grey box system identification, and the parameters were validated by fitting the reconstructed model to a set of flight data with a different excitation waveform. The flight controller with model predictive control was formed using the identified models for flight simulation. The results demonstrate that the system identification results are able to provide reference models for the model-based controller development of the novel-design tail-sitter UAV.

Impact of Direct Tax and Indirect Tax on Economic Growth in Vietnam

Tax can be categorised into direct tax and indirect tax. This paper uses the ordinary least-squares regression method to study the impact of direct and indirect tax on economic growth in Vietnam in the period 2003-2017. Statistical data is collected from the Ministry of Finance of Vietnam. Theoretically, tax generates the state budget revenue and is a tool to regulate the economy. The results of statistical tests show that tax has a positive impact on Vietnam's economic growth. However, the effects of direct tax and indirect tax are different. The indirect tax has a positive influence and promote Vietnam's economic growth, while the impact of the direct tax is invisible. There has not been sufficient evidence to confirm that the indirect tax has a more positive impact than the direct tax. To promote economic growth, Vietnam needs to restructure its tax system towards: (1) Increasing the proportion of indirect tax, reducing the proportion of direct tax in the state budget revenue; (2) Expanding tax bases; (3) Reducing tax rates of corporate income tax and personal income tax; (4) Increasing tax rates of environmental protection tax, natural resources tax, value added tax and excise tax on some types of goods which harm health and environment.

An Empirical Investigation on the Relation between Disclosure and Financial Performance of Islamic Banks in the United Arab Emirates

The paper examines the level of disclosure on Islamic banks' performance in the United Arab Emirates (UAE). The data was collected through content analysis of annual reports and financial statements of all fully-fledged Islamic banks working in the UAE over the period 2009 to 2013. Return on Assets is used as a proxy for the performance of Islamic banks while disclosure index is used as a proxy for Islamic banks' disclosure. Also, predetermined variables are used in the study like Size, Deposits, Non-Performing Investments and Capital to Risk Weighted Assets Ratio. Two-Stage Least-Square regression method is used to check the interdependence relationships between disclosure and performance of Islamic banks in the UAE. The results show a significant relationship between performance and disclosure in the UAE Islamic banks. Our regression results show that Islamic banks with higher levels of disclosure lead to higher operating performance. Furthermore, the performance has a great impact on the level of disclosure which means Islamic banks with high performance measures will disclose more information for investors and other institutions in order to reduce the cost of equity and increase their values in the market. This study is considered as a battery for further studies in the relationship between disclosure and financial performance of Islamic banks at a global level.

Effect of Eating Breakfast on Cognitive Development of Elementary and Middle School Students: An Empirical Study Using Large-Scale Provincial Survey Data.

BackgroundIn existing research on the importance of student breakfast, few studies have focused on the impact of breakfast on student cognitive development. Further, empirical research in this field is mostly based correlation analysis, in which it is difficult to control the influence of selection bias on the analysis results.Material/MethodsHere, we used student academic performance, based on the academic quality monitoring data of Jiangsu basic education students, as a proxy variable for cognitive development, and used both ordinary least-squares regression and propensity score matching methods to analyze the impact of eating breakfast on the cognitive development of primary and middle school students.ResultsWe found that it is still common for students in primary and secondary schools to go without breakfast, and that this is even true in middle schools. Whether students eat breakfast is affected by many factors, and the frequency of eating breakfast has a significant positive impact on student achievement.ConclusionsIn primary school, students who eat breakfast every day in a week scored 31.322 points higher in academic performance than those who did not. In middle school, students who ate breakfast on time every day had significantly better academic performance (31.335 points higher) than those who did not eat breakfast every day. This indicates that eating breakfast every day has a significant effect on the cognitive development of students.

Feasibility of Using Rice Leaves Hyperspectral Data to Estimate CaCl2-extractable Concentrations of Heavy Metals in Agricultural Soil

Heavy metals contamination is a serious problem of China. It is necessary to estimate bioavailability concentrations of heavy metals in agricultural soil for keeping the food security and human health. This study aimed to use hyperspectral data of rice (Oryza sativa) leaves as an indicator to retrieve the CaCl2-extractable concentrations of heavy metals in agricultural soil. Twenty-one rice samples, soil samples and reflectance spectra of rice leaves were collected, respectively. The potential relations between hyperspectral data and CaCl2-extractable heavy metals (E-HM) were explored. The partial least-squares regression (PLSR) method with leave-one-out cross-validation has been used to predict concentrations of CaCl2-extractable cadmium (E-Cd) and concentrations of CaCl2-extractable lead (E-Pb) in farmland soil. The results showed that the concentrations of E-Cd in soil had significant correlation with concentrations of Cd in rice leaves; the number of bands associated with E-Cd was more than that of E-Pb. Four indices (normalized difference vegetation index (NDVI), carotenoid reflectance index (CRI), photochemical reflectance index 2 (PRI2), normalized pigments chlorophyll ratio index (NPCI)) were significant (P < 0.05) and negatively related to the E-Cd concentrations. The PLSR model of E-Cd concentrations performed better than the PLSR model of E-Pb concentrations, which with R2 = 0.592 and RMSE = 0.046. We conclude that if the rice was sensitive to E-HM and/or the crop was stressed by the E-HM, the hyperspectral data of field rice leaves hold potentials in estimating concentration of E-HM in farmland soil. Therefore, this method provides a new insight to monitoring the E-HM content in agricultural soil.

Quantitative Structure-activity Relationship Analysis for Predicting Lipophilicity of Aniline Derivatives (Including some Pharmaceutical Compounds).

In this study, we used a hierarchical approach to develop quantitative structureactivity relationship (QSAR) models for modeling lipophilicity of a set of 81 aniline derivatives containing some pharmaceutical compounds. The multiple linear regression (MLR), principal component regression (PCR) and partial least square regression (PLSR) methods were utilized to construct QSAR models. Quantum mechanical calculations at the density functional theory level and 6- 311++G** basis set were carried out to obtain the optimized geometry and then, the comprehensive set of molecular descriptors was computed by using the Dragon software. Genetic algorithm (GA) was applied to select suitable descriptors which have the most correlation with lipophilicity of the studied compounds. It was identified that such descriptors as Barysz matrix (SEigZ), hydrophilicity factor (Hy), Moriguchi octanol-water partition coefficient (MLOGP), electrophilicity (ω/eV) van der Waals volume (vWV) and lethal concentration (LC50/molkg-1) are the best descriptors for QSAR modeling. The high correlation coefficients and the low prediction errors for MLR, PCR and PLSR methods confirmed good predictability of the three models. In present study, the high correlation between experimental and predicted logP values of aniline derivatives indicated the validation and the good quality of the resulting three regression methods, but MLR regression procedure was a little better than the PCR and PLSR methods. It was concluded that the studied aniline derivatives are not hydrophilic compounds and this means these compounds hardly dissolve in water or an aqueous solvent.

Biopolymer Drill-in Fluid Performance for Different Rheological Models using Statistical Characterisation

Appropriate selection of rheological models is important for hydraulic calculations of pressure loss prediction and hole cleaning efficiency of drilling fluids. Power law, Bingham-Plastic and Herschel-Bulkley models are the conventional fluid models used in the oilfield. However, there are other models that have been proposed in literature which are under/or not utilized in the petroleum industry. The primary objective of this paper is to recommend a rheological model that best-fits the rheological behaviour of xanthan gum-based biopolymer drill-in fluids for hydraulic evaluations. Ten rheological models were evaluated in this study. These rheological models have been posed deterministically and due to the unrealistic nature have been replaced by statistical models, by adding an error (disturbance) term and making suitable assumptions about them. Rheological model parameters were estimated by least-square regression method. Models like Sisko and modified Sisko which are not conventional models in oil industry gave a good fit. Modified Sisko model which is a four parameter rheological model was selected as the best-fit model since it produced the least residual mean square of 0.61 Ibf2/100ft4. There is 95% certainty that the true best-fit curve lies within the confidence band of this function of interest. Keywords: Biopolymer; Least-Square Regression; Residual Mean Squares; Rheologram

Towards an optical multisensor system for dairy: Global calibration for fat analysis in homogenized milk

A new approach to constructing global calibration models for optical spectroscopic determination of fat content in normalized milk (homogenized milk with standardized nutrient content) in the region 400–1100 nm has been developed. Following the global modelling approach attempting to cover all possible variability of analyzed object, calibration models were built on a large set of commercial milk samples from many producers, collected in two different countries and varying from each other in nutrient composition and methods of technological processing. Suggested method employs spectral information related to the light scatter effect by colloidal particles of the milk constituents. Due to the application of advanced variable interval selection method, the root mean-square error of cross-validation (RMSECV) value of 0.12% in the range 1.5–6.0% of fat content has been achieved in the calibration by partial least-squares (PLS) regression method. Selected spectral intervals can be used for the development of an inexpensive and compact optical sensor for milk analysis based on light-emitting diodes (LEDs). Global calibration models developed in this study can be used for the practical analysis of fat in various normalized milk samples.

Multi-model ensemble prediction model for carbon efficiency with application to iron ore sintering process

Iron ore sintering is one of the most energy-consuming process in steel industry. Accurate prediction of carbon efficiency for this process is beneficial to energy savings and consumption reduction. Considering the sintering process exhibits strong nonlinearities, multiple parameters, multiple operating conditions, etc., a multi-model ensemble prediction model based on the actual run data is developed to achieve the high-precision prediction of carbon efficiency. It takes the comprehensive coke ratio (CCR) as a metric (index) of carbon efficiency in the sintering process. First, an affinity propagation clustering algorithm is used to realize the automatic identification of multiple operating conditions. Then, different models are established under different operating conditions by using the proposed least squares support vector machine (LS-SVM) with hybrid kernel modeling method. Finally, a partial least-squares regression method is employed as an ensemble strategy to combine the different models to form the multi-model ensemble prediction model for the CCR. The simulation results involving the actual run data demonstrate that the proposed model can predict the CCR accurately when compared with other prediction methods. The results of actual runs show that the coefficient of determination for the proposed model is 0.877. The proposed model satisfies the requirements of actual sintering process and enables the real-time prediction.

Nonlinear Estimation of Variables for Heat Release Calculation Using a Gradient Method

Advanced combustion such as the partially premixed combustion (PPC) is characterized by high energy efficiency. However, they are sensitive to the inlet condition and injection of a combustion engine. Therefore, it is essential to use combustion feedback. The accuracy of the feedback variables, derived from the cylinder pressure signal, is crucial for effective combustion feedback control. This paper proposes a nonlinear least-squares regression method to estimate the pressure offsets and variable polytropic exponent in heat release calculation automatically. The combustion feedback variables derived from the auto-tuned heat-release rate were applied to a heavy-duty compression ignition (CI) engine burning with gasoline fuel.

Effects of Growth Stage Development on Paddy Rice Leaf Area Index Prediction Models

A in situ hyperspectral dataset containing multiple growth stages over multiple growing seasons was used to build paddy rice leaf area index (LAI) estimation models with a special focus on the effects of paddy rice growth stage development. The univariate regression method applied to the vegetation index (VI), the traditional multivariate calibration method of partial least squares regression (PLSR), and modern machine learning methods such as support vector regression (SVR), random forests (RF), and artificial neural networks (ANN) based on the original and first-derivative hyperspectral data were evaluated in this study for paddy rice LAI estimation. All the models were built on the whole growing season and on each separate vegetative, reproductive and ripening growth stage of paddy rice separately. To ensure a fair comparison, the models of the whole growing season were also validated on data for each separate growth stage of the standalone validation dataset. Moreover, the optimal band pairs for calculating narrowband difference vegetative index (DVI), normalized difference vegetation index (NDVI) and simple ratio vegetation index (SR) were determined for the whole growing season and for each separate growth stage separately. The results showed that for both the whole growing season and for each single growth stage, the red-edge and near-infrared band pairs are optimal for formulating the narrowband DVI, NDVI and SR. Among the four multivariate calibration methods, SVR and RF yielded more accurate results than the other two methods. The SVR and RF models built on first-derivative spectra provided more accurate results than the corresponding models on the original spectra for both whole growing season models and separate growth stage models. Comparing the prediction accuracy based on the whole growing season revealed that the RF and SVR models showed an advantage over the VI models. However, comparing the prediction accuracy based on each growth stage separately showed that the VI models provided more accurate results for the vegetative growth stages. The SVR and RF models provided more accurate results for the ripening growth stage. However, the whole growing season RF model on first-derivative spectra could provide reasonable accuracy for each single growth stage.

The mixture problem in flake analysis: Allocating flake samples to segments of reduction using CLSR methods

Flake assemblages often are mixtures from knapping episodes that vary by reduction mode or stage. Stahle and Dunn (1982) developed a constrained least-squares regression (CLSR) method to allocate hypothetically mixed chert assemblages to successive biface-reduction stages, a partial solution to the mixing problem. ConReg is a recently developed alternative, validated against Stahle and Dunn's and our own experimental controls from replication of obsidian biface preforms. Using both controls, we apply ConReg to the “unmixing” of flake assemblages from North American Great Basin obsidian quarries. Controls produce different solutions; their separate comparison to independent analytical results of stage composition favor ConReg used with our controls. What appeared to be relatively similar early “stage” assemblages are resolved to finer proportional allocation across wider ranges of the reduction continuum, in the process showing considerable variation between samples that otherwise might escape detection. The mixture problem is a challenge to all approaches to flake analysis, not just mass analysis, but one that ConReg can help us surmount.

Refined video segmentation through global appearance regression

To achieve accurate segmentation in unconstrained videos, we propose a novel segmentation framework based on a two-stream deep convolution network. Our algorithm exploits the object’s robust pixel-level features within all the video frames and generates foreground likelihood maps with sufficient details. At first, a two-stream video segmentation network using multiple hierarchical features is designed to generate initial segmentation masks. Then, all initial segmentation masks and original corresponding images are collected to learn an appearance model by the least-square regression method. The model computes appearance likelihood map for every image. Finally, pairs of initial segmentation masks and appearance likelihood maps are fused by a proposed fusion network to generate final high-quality segmentation maps. Experiments on the challenging dataset DAVIS verify the effectiveness of our appearance regression and demonstrate that our proposed algorithm outperforms the state-of-the-art algorithms.

Intelligent Multisensor System For Analytical Control Of Sausages

The new technique of intelligent analysis of chemical aroma patterns of boiled sausages obtained by the electronic nose for authentication and microbiological safety assessment is developed. The informativeness of features extracted from steady-state responses of the multisensor system and robustness of chemometric algorithms for solving the objectives of qualitative and quantitative analysis of sausage volatile compounds are investigated. The classification model was built using maximum response values as input vectors of an optimized probabilistic neural network, which allows obtaining a 100 % accuracy of different sample grades identification and detection samples adulterated with soy protein. The method of partial least squares regression and area values as features were used for regression modelling and prediction of QMAFAnM with a relative error less than 12 % for a microbiological safety assessment of previously identified sausages. The use of the robust analytical technique to assess authentication, adulteration, total bacterial count for one measurement using the electronic nose in combination with machine learning algorithms will allow to significantly reduce the measurement time and the cost of analysis, and avoid subjective estimation of the results.

Optimizing process of preparing artificial-similar material for rocky slope with uniform formula design

The determination of material formula needs try-and-error experiment, and consumes large amount of time and fund. In order to solve the problem, a comprehensive method is established, via the experiment of artificial-similar material formula of a mine slope. We controlled the samples by the compactness, and arranged the formula of the test group with the method of the uniform formula experiment. The physical and mechanical parameters of these samples were analyzed using the method of the partial least-squares regression (PLS). And a mathematical model of the indexes of physical and mechanics parameters relating to the factors of formulation constituents was established eventually. We used the model to analyze the effect of each formulation constituent on physical and mechanics parameters of samples. The experiment results and analysis illustrates that 1) in the formulation of similar material, the effect of raw materials on the internal friction angle φ and cohesion C is opposite; 2) The method can highly facilitate the process of the of preparing artificial-similar materials, more economic and effective.

Partial Least Squares Identification of Multi Look-Up Table Digital Predistorters for Concurrent Dual-Band Envelope Tracking Power Amplifiers

This paper presents a technique to estimate the coefficients of a multiple-look-up table (LUT) digital predistortion (DPD) architecture based on the partial least-squares (PLS) regression method. The proposed 3-D distributed memory LUT architecture is suitable for efficient FPGA implementation and compensates for the distortion arising in concurrent dual-band envelope tracking power amplifiers. On the one hand, a new variant of the orthogonal matching pursuit algorithm is proposed to properly select only the best LUTs of the DPD function in the forward path, and thus reduce the number of required coefficients. On the other hand, the PLS regression method is proposed to address both the regularization problem of the coefficient estimation and, at the same time, reducing the number of coefficients to be estimated in the DPD feedback identification path. Moreover, by exploiting the orthogonality of the PLS transformed matrix, the computational complexity of the parameters’ identification can be significantly simplified. Experimental results will prove how it is possible to reduce the DPD complexity (i.e., the number of coefficients) in both the forward and feedback paths while meeting the targeted linearity levels.

Numerical Simulation of Propagation Characteristics of Hazardous Noxious Substances Spilled from Transport Ships

This study numerically investigates the propagation characteristics of hazardous noxious substances (HNSs) spilled from transport ships and suggests the metal model for predicting the HNS propagation velocity varied with the current velocity and HNS density. The commercial computational fluid dynamics (CFD) code ANSYS FLUENT (V. 17.2) was used for two-dimensional simulation based on the Reynolds-averaged Navier–Stokes (RANS) equation together with the standard k–ε model. The scalar transport equation was also solved to estimate the spatial and transient behaviors of HNS. The main parameters to analyze the near-field propagation characteristics of HNSs spilled from the ship were layer thickness, HNS concentration, and propagation velocity. It was found that advection becomes more dominant in propagating an HNS layer that becomes thinner as the current velocity increases. When the current velocity increased beyond a certain level (~0.75 m/s), the mixing effect made the HNS layer less dense but thicker. Consequently, lower-density HNS causes increased HNS concentrations at sea level. As the current velocity increased, the concentration distribution became homogeneous regardless of HNS density. In particular, the second-order response surface model provided for three variables on the basis of the numerical results for 15 cases with the use of the general least-squares regression method, showing a good fit. This model would be useful in estimating the propagation velocity of HNS spilled from a ship.

Evaluation of Carbohydrate Concentrations in Phalaenopsis Using Near-infrared Spectroscopy

Carbohydrate concentrations are important indicators of the internal quality of Phalaenopsis. In this study, near-infrared (NIR) spectroscopy was used for quantitative analyses of fructose, glucose, sucrose, and starch in Phalaenopsis plants. Both modified partial least-squares regression (MPLSR) and stepwise multiple linear regression (SMLR) methods were used for spectral analysis of 302 Phalaenopsis samples in the full visible NIR wavelength range (400–2498 nm). Calibration models built by MPLSR were better than those built by SMLR. For fructose, the smoothed first derivative MPLSR model provided the best results, with a correlation coefficient of calibration (Rc) of 0.96, standard error of calibration (SEC) of 0.22% dry weight (DW), standard error of validation (SEV) of 0.28% DW, and bias of -0.01% DW. For glucose, the MPLSR model based on the smoothed first derivative spectra was the best (Rc = 0.96; SEC = 0.26% DW; SEV = 0.32% DW; and bias = 0.01% DW). The best MPLSR model of sucrose was developed using the smoothed first derivative spectra (Rc = 0.96; SEC = 0.24% DW; SEV = 0.31% DW; bias = -0.03% DW). Regarding starch, the smoothed first derivative MPLSR model showed the best effects (Rc = 0.91; SEC = 0.47% DW; SEV = 0.56% DW; bias = -0.02% DW). Both the MPLSR and SMLR models showed satisfactory predictability, indicating that NIR has the potential to be adopted as an effective method of rapid and accurate inspection of the carbohydrate concentrations of Phalaenopsis plants. This technique could contribute substantially to quality management of Phalaenopsis.