Intelligent Multivariable Modeling of Blast Furnace Molten Iron Quality Based on Dynamic AGA-ANN and PCA

Blast furnace （BF） ironmaking process has complex and nonlinear dynamic characteristics. The molten iron temperature （MIT） as well as Si, P and S contents of molten iron is difficult to be directly measured online, and large-time delay exists in offline analysis through laboratory sampling. A nonlinear multivariate intelligent modeling method was proposed for molten iron quality （MIQ） based on principal component analysis （PCA） and dynamic ge- netic neural network. The modeling method used the practical data processed by PCA dimension reduction as inputs of the dynamic artificial neural network （ANN）. A dynamic feedback link was introduced to produce a dynamic neu- ral network on the basis of traditional back propagation ANN. The proposed model improved the dynamic adaptabili- ty of networks and solved the strong fluctuation and resistance problem in a nonlinear dynamic system. Moreover, a new hybrid training method was presented where adaptive genetic algorithms （AGA） and ANN were integrated, which could improve network convergence speed and avoid network into local minima. The proposed method made it easier for operators to understand the inside status of blast furnace and offered real-time and reliable feedback infor- mation for realizing close-loop control for MIQ. Industrial experiments were made through the proposed model based on data collected from a practical steel company. The accuracy could meet the requirements of actual operation.

Fractional derivative multivariable grey model for nonstationary sequence and its application

Most of the existing multivariable grey models are based on the 1-order derivative and 1-order accumulation, which makes the parameters unable to be adjusted according to the data characteristics of the actual problems. The results about fractional derivative multivariable grey models are very few at present. In this paper, a multivariable Caputo fractional derivative grey model with convolution integral CFGMC(q, N) is proposed. First, the Caputo fractional difference is used to discretize the model, and the least square method is used to solve the parameters. The orders of accumulations and differential equations are determined by using particle swarm optimization(PSO). Then, the analytical solution of the model is obtained by using the Laplace transform, and the convergence and divergence of series in analytical solutions are also discussed. Finally, the CFGMC(q, N) model is used to predict the municipal solid waste(MSW). Compared with other competition models, the model has the best prediction effect. This study enriches the model form of the multivariable grey model, expands the scope of application, and provides a new idea for the development of fractional derivative grey model.

Designing Artemisinins with Antimalarial Potential, Combining Molecular Electrostatic Potential, Ligand-Heme Interaction and Multivariate Models

Artemisinins tested against W-2 strains of malaria falciparum are investigated with molecular electrostatic potential (MEP), in an attempt to identify key features of the compounds that are necessary for their activities, as well as to investigate likely interactions with the receptor in a biological process and to use that information to propose new molecules. In order to discover the best geometry involving the ligand-receptor complexes (heme) studied and help in the proposition of the new derivatives, molecular simulations of interactions between the most negative charged region around the peroxide and heme locates (the ones around the Fe2+ ion) were carried out. In addition, PCA (principal components analysis), HCA (hierarchical cluster analysis), SDA (stepwise discriminant analysis), and KNN (K-nearest neighbor) multivariate models were employed to investigate which descriptors are responsible for the classification between the higher and lower antimalarial activity of the compounds, and also this information was used to propose new potentially active molecules. The information accumulated in studies of MEP, molecular docking, and multivariate analysis supported the proposal of new structures with potential antimalarial activities. The multivariate models constructed were applied to the new structures and indicated numbers 19 and 20 as the most prominent for syntheses and biological assays.

Multivariate predictive model for asymptomatic spontaneous bacterial peritonitis in patients with liver cirrhosis

BACKGROUNDSpontaneous bacterial peritonitis (SBP) is a detrimental infection of the asciticfluid in liver cirrhosis patients, with high mortality and morbidity. Earlydiagnosis and timely antibiotic administration have successfully decreased themortality rate to 20%-25%. However, many patients cannot be diagnosed in theearly stages due to the absence of classical SBP symptoms. Early diagnosis ofasymptomatic SBP remains a great challenge in the clinic.AIMTo establish a multivariate predictive model for early diagnosis of asymptomaticSBP using positive microbial cultures from liver cirrhosis patients with ascites.METHODSA total of 98 asymptomatic SBP patients and 98 ascites liver cirrhosis patients withnegative microbial cultures were included in the case and control groups,respectively. Multiple linear stepwise regression analysis was performed toidentify potential indicators for asymptomatic SBP diagnosis. The diagnosticperformance of the model was estimated using the receiver operatingcharacteristic curve.RESULTSPatients in the case group were more likely to have advanced disease stages,cirrhosis related-complications, worsened hematology and ascites, and higher mortality. Based on multivariate analysis, the predictive model was as follows: y (P) = 0.018 + 0.312 × MELD (model of end-stage liver disease) + 0.263 × PMN(ascites polymorphonuclear) + 0.184 × N (blood neutrophil percentage) + 0.233 ×HCC (hepatocellular carcinoma) + 0.189 × renal dysfunction. The area under thecurve value of the established model was 0.872, revealing its high diagnosticpotential. The diagnostic sensitivity was 73.5% (72/98), the specificity was 86.7%(85/98), and the diagnostic efficacy was 80.1%.CONCLUSIONOur predictive model is based on the MELD score, polymorphonuclear cells,blood N, hepatocellular carcinoma, and renal dysfunction. This model mayimprove the early diagnosis of asymptomatic SBP.

Artificial neural network models predicting the leaf area index:a case study in pure even-aged Crimean pine forests from Turkey

Background: Leaf Area Index(LAI) is an important parameter used in monitoring and modeling of forest ecosystems. The aim of this study was to evaluate performance of the artificial neural network(ANN) models to predict the LAI by comparing the regression analysis models as the classical method in these pure and even-aged Crimean pine forest stands.Methods: One hundred eight temporary sample plots were collected from Crimean pine forest stands to estimate stand parameters. Each sample plot was imaged with hemispherical photographs to detect the LAI. The partial correlation analysis was used to assess the relationships between the stand LAI values and stand parameters, and the multivariate linear regression analysis was used to predict the LAI from stand parameters. Different artificial neural network models comprising different number of neuron and transfer functions were trained and used to predict the LAI of forest stands.Results: The correlation coefficients between LAI and stand parameters(stand number of trees, basal area, the quadratic mean diameter, stand density and stand age) were significant at the level of 0.01. The stand age, number of trees, site index, and basal area were independent parameters in the most successful regression model predicted LAI values using stand parameters(R_(adj)~2=0.5431). As corresponding method to predict the interactions between the stand LAI values and stand parameters, the neural network architecture based on the RBF 4-19-1 with Gaussian activation function in hidden layer and the identity activation function in output layer performed better in predicting LAI(SSE(12.1040), MSE(0.1223), RMSE(0.3497), AIC(0.1040), BIC(-77.7310) and R^2(0.6392)) compared to the other studied techniques.Conclusion: The ANN outperformed the multivariate regression techniques in predicting LAI from stand parameters. The ANN models, developed in this study, may aid in making forest management planning in study forest stands.

Improving reservoir volumetric estimations in petroleum resource assessment using discovery process models

The reservoir volumetric approach represents a widely accepted, but flawed method of petroleum play resource calculation. In this paper, we propose a combination of techniques that can improve the applicability and quality of the resource estimation. These techniques include： 1） the use of the Multivariate Discovery Process model （MDP） to derive unbiased distribution parameters of reservoir volumetric variables and to reveal correlations among the variables; 2） the use of the Geo-anchored method to estimate simultaneously the number of oil and gas pools in the same play; and 3） the crossvalidation of assessment results from different methods. These techniques are illustrated by using an example of crude oil and natural gas resource assessment of the Sverdrup Basin, Canadian Archipelago. The example shows that when direct volumetric measurements of the untested prospects are not available, the MDP model can help derive unbiased estimates of the distribution parameters by using information from the discovered oil and gas accumulations. It also shows that an estimation of the number of oil and gas accumulations and associated size ranges from a discovery process model can provide an alternative and efficient approach when inadequate geological data hinder the estimation. Cross-examination of assessment results derived using different methods allows one to focus on and analyze the causes for the major differences, thus providing a more reliable assessment outcome.

All Admissible Linear Estimators under Quadratic Loss in Multivariate Model

For multivariate linear model Y=XΘ+ε, ~N(0, σ 2ΣV), this paper is concerned with the admissibility of linear estimators of estimable function SXΘ in the class of all estimators. All admissible linear estimators of SXΘ are given under each of four definitions of admissibility.

Joint multivariate statistical model and its applications to the synthetic earthquake prediction

Considering the problems that should be solved in the synthetic earthquake prediction at present, a new model is proposed in the paper. It is called joint multivariate statistical model combined by principal component analysis with discriminatory analysis. Principal component analysis and discriminatory analysis are very important theories in multivariate statistical analysis that has developed quickly in the late thirty years. By means of maximization information method, we choose several earthquake prediction factors whose cumulative proportions of total sam-ple variances are beyond 90% from numerous earthquake prediction factors. The paper applies regression analysis and Mahalanobis discrimination to extrapolating synthetic prediction. Furthermore, we use this model to charac-terize and predict earthquakes in North China (30~42N, 108~125E) and better prediction results are obtained.

THE COMPRESSION LS ESTIMATE OF REGRESSION COEFFICIENT IN MULTIVARIATE LINEAR MODEL

In this paper, compression LS estimate (k) of the regression coefficient B isconsidered when the design matrix present ill-condition in multivariate linear model.The MSE (mean square error)of the estimate(k)=Vec( (k))is less than theMSE of LS estimate β￣* of the regression coefficient β= Vec(B) by choosing the pa-rameter k. Admissibility , numerical stability and relative efficiency of (k)are proved. The method of determining k value for practical use is also suggested

Dynamic Hedging Based on Markov Regime-Switching Dynamic Correlation Multivariate Stochastic Volatility Model

It is important to consider the changing states in hedging.The Markov regime-switching dynamic correlation multivariate stochastic volatility( MRS-DC-MSV) model was proposed to solve this issue. DC-MSV model and MRS-DC-MSV model were used to calculate the time-varying hedging ratios and compare the hedging performance. The Markov chain Monte Carlo( MCMC) method was used to estimate the parameters. The results showed that,there were obviously two economic states in Chinese financial market. Two models all did well in hedging,but the performance of MRS-DCMSV model was better. It could reduce risk by nearly 90%. Thus,in the hedging period,changing states is a factor that cannot be neglected.

Epidemic Prediction of Hemorrhagic Fever with Renal Syndrome Using Grey Multivariable Model

In this paper,we set up a grey multivtriable model to predict the HFRS morbidity.Forecasting test results show that this new model is of general use in the prediction Of the disease.It is particularly appropriate for HFRS investigations and controls.

Diagnostic value evaluation of trefoil factors family 3 for the early detection of colorectal cancer

AIM The purpose of this study was to evaluate the diagnostic value of trefoil factor family 3(TFF3) for the early detection of colorectal cancer(CC). METHODS Serum TFF3 and carcino-embryonic antigen(CEA) were detected in 527 individuals, including 115 healthy control(HC), 198 colorectal adenoma(CA), and 214 CC individuals in the training group. RESULTS Serum TFF3 showed no significant correlation with age, gender, or tumor location but showed significant correlation with the tumor stage. Serum TFF3 in the CC group was significantly higher than in the HC or CA group. The AUC values of TFF3 for discriminating between HC and CC and between CA and CC were 0.930(0.903, 0.958) and 0.834(0.796, 0.873). A multivariate model combining TFF3 and CEA was built. Compared to TFF3 or CEA alone, the multivariate model showed significant improvement(P < 0.001). For discriminating between HC and CC, HC and early stage CC, HC and advanced stage CC, CA and CC, CA and early stage CC, and CA and advanced stage CC in the training group, the sensitivities were 92.99%, 91.46%, 93.18%, 73.83%, 76.83%, and 81.82%, and the specificities were 91.30%, 91.30%, 93.91%, 88.38%, 77.27%, and 88.38%, respectively. After validation, the sensitivities were 89.39%, 85.71%, 90.79%, 72.73%, 71.43%, and 78.95%, and the specificities were 87.85%, 87.85%, 2.52%, 87.85%, 80.77%, and 87.50%, respectively. CONCLUSION The multivariate diagnostic model that included TFF3 and CEA showed significant improvement over the conventional biomarker CEA and might provide a potential method for the early detection of CC.

Studies on Heavy Metal Pollution in Soil-Plant System:A Review

Heavy metal pollution in soil-plant system is of major environmental concern on a world scale and in China in particular with the rapid development of industry. The heavy metal pollution status in soil-plant system in China, the research progress on the bioavailability of heavy metals (affecting factors, extraction methods, free-ion activity model, adsorption model, multivariate regression model, Q-I relationship, and compound pollution), and soil remediation are reviewed in the paper. Future research and monitoring is also discussed.

SOME PROPERTIES OF ROSEN’S MLE FOR GENERAL DISTRIBUTIONS

Von Rosen (1989) proposed the MLE of parameters in multivariate linear normal model MLNM(sumfromn= lto ∞AiBiCi). This paper discusses some properties of Rosen's MLE for general distributions which includs invariant, equivariant, strong consistency and asymptotic normality. Furthermore, we can construct the consistent confidence region for the parameter of experctation in MLNM(sumfromn=1to∞, AiBiCi) and obtain asymptotic distribu- tion and consistent confidence region of the linear discrimination function for canonical correlation by Kahtri (1988).

Predictability of well construction time with multivariate probabilistic approach

Current univariate approach to predict the probability of well construction time has limited accuracy due to the fact that it ignores key factors affecting the time.In this study,we propose a multivariate probabilistic approach to predict the risks of well construction time.It takes advantage of an extended multi-dimensional Bernacchia–Pigolotti kernel density estimation technique and combines probability distributions by means of Monte-Carlo simulations to establish a depth-dependent probabilistic model.This method is applied to predict the durations of drilling phases of 192 wells,most of which are located in the AustraliaAsia region.Despite the challenge of gappy records,our model shows an excellent statistical agreement with the observed data.Our results suggested that the total time is longer than the trouble-free time by at least 4 days,and at most 12 days within the 10%–90% confidence interval.This model allows us to derive the likelihoods of duration for each phase at a certain depth and to generate inputs for training data-driven models,facilitating evaluation and prediction of the risks of an entire drilling operation.

Data Fusion about Serviceability Reliability Prediction for the Long-Span Bridge Girder Based on MBDLM and Gaussian Copula Technique

This article presented a new data fusion approach for reasonably predicting dynamic serviceability reliability of the long-span bridge girder.Firstly,multivariate Bayesian dynamic linear model(MBDLM)considering dynamic correlation among the multiple variables is provided to predict dynamic extreme deflections;secondly,with the proposed MBDLM,the dynamic correlation coefficients between any two performance functions can be predicted;finally,based on MBDLM and Gaussian copula technique,a new data fusion method is given to predict the serviceability reliability of the long-span bridge girder,and the monitoring extreme deflection data from an actual bridge is provided to illustrated the feasibility and application of the proposed method.

Some Convergence Properties for Weighted Sums of Martingale Difference Random Vectors

Let{X_(ni),F_(ni);1≤i≤n,n≥1}be an array of R^(d)martingale difference random vectors and{A_(ni),1≤i≤n,n≥1}be an array of m×d matrices of real numbers.In this paper,the Marcinkiewicz-Zygmund type weak law of large numbers for maximal weighted sums of martingale difference random vectors is obtained with not necessarily finite p-th(1<p<2)moments.Moreover,the complete convergence and strong law of large numbers are established under some mild conditions.An application to multivariate simple linear regression model is also provided.

The Superiorities of Bayes Linear Unbiased Estimator in Multivariate Linear Models

In this article, the Bayes linear unbiased estimator （BALUE） of parameters is derived for the multivariate linear models. The superiorities of the BALUE over the least square estimator （LSE） is studied in terms of the mean square error matrix （MSEM） criterion and Bayesian Pitman closeness （PC） criterion.

pyvine:The Python Package for Regular Vine Copula Modeling,Sampling and Testing

Regular vine copula provides rich models for dependence structure modeling.It combines vine structures and families of bivariate copulas to construct a number of multivariate distributions that can model a wide range dependence patterns with different tail dependence for different pairs.Two special cases of regular vine copulas,C-vine and D-vine copulas,have been extensively investigated in the literature.We propose the Python package,pyvine,for modeling,sampling and testing a more generalized regular vine copula(R-vine for short).R-vine modeling algorithm searches for the R-vine structure which maximizes the vine tree dependence in a sequential way.The maximum likelihood estimation algorithm takes the sequential estimations as initial values and uses L-BFGS-B algorithm for the likelihood value optimization.R-vine sampling algorithm traverses all edges of the vine structure from the last tree in a recursive way and generates the marginal samples on each edge according to some nested conditions.Goodness-of-fit testing algorithm first generates Rosenblatt’s transformed data E and then tests the hypothesis H^(∗)_(0):E∼C_(⊥)by using Anderson–Darling statistic,where C_(⊥)is the independence copula.Bootstrap method is used to compute an adjusted p-value of the empirical distribution of replications of Anderson–Darling statistic.The computing of related functions of copulas such as cumulative distribution functions,Hfunctions and inverse H-functions often meets with the problem of overflow.We solve this problem by reinvestigating the following six families of bivariate copulas:Normal,Student t,Clayton,Gumbel,Frank and Joe’s copulas.Approximations of the above related functions of copulas are given when the overflow occurs in the computation.All these are implemented in a subpackage bvcopula,in which subroutines are written in Fortran and wrapped into Python and,hence,good performance is guaranteed.

Empirical likelihood estimation in multivariate mixture models with repeated measurements

Multivariate mixtures are encountered in situations where the data are repeated or clustered measurements in the presence of heterogeneity among the observations with unknown proportions.In such situations,the main interest may be not only in estimating the component parameters,but also in obtaining reliable estimates of the mixing proportions.In this paper,we propose an empirical likelihood approach combined with a novel dimension reduction procedure for estimating parameters of a two-component multivariate mixture model.The performance of the new method is compared to fully parametric as well as almost nonparametric methods used in the literature.