AFSTGCN:Prediction for multivariate time series using an adaptive fused spatial-temporal graph convolutional network

The prediction for Multivariate Time Series(MTS)explores the interrelationships among variables at historical moments,extracts their relevant characteristics,and is widely used in finance,weather,complex industries and other fields.Furthermore,it is important to construct a digital twin system.However,existing methods do not take full advantage of the potential properties of variables,which results in poor predicted accuracy.In this paper,we propose the Adaptive Fused Spatial-Temporal Graph Convolutional Network(AFSTGCN).First,to address the problem of the unknown spatial-temporal structure,we construct the Adaptive Fused Spatial-Temporal Graph(AFSTG)layer.Specifically,we fuse the spatial-temporal graph based on the interrelationship of spatial graphs.Simultaneously,we construct the adaptive adjacency matrix of the spatial-temporal graph using node embedding methods.Subsequently,to overcome the insufficient extraction of disordered correlation features,we construct the Adaptive Fused Spatial-Temporal Graph Convolutional(AFSTGC)module.The module forces the reordering of disordered temporal,spatial and spatial-temporal dependencies into rule-like data.AFSTGCN dynamically and synchronously acquires potential temporal,spatial and spatial-temporal correlations,thereby fully extracting rich hierarchical feature information to enhance the predicted accuracy.Experiments on different types of MTS datasets demonstrate that the model achieves state-of-the-art single-step and multi-step performance compared with eight other deep learning models.

Multivariable sales prediction for filling stations via GA improved BiLSTM

Accurate sales prediction in filling stations is the basis to fill in the refined oil in time and avoid the outof-stock as much as possible.Considering the defect of great“lag”in the general time series model,this paper summarizes the multiple factors that influence the oil sales and develops a multivariable long short-term memory(LSTM)neural network,with the hyper-parameters being improved by the genetic algorithm(GA).To further improve the prediction accuracy,the proposed LSTM neural network is generalized to bidirectional LSTM(Bi LSTM),in which the impact of future factors on present sales can be taken into account by backward training.Finally,different LSTM structures and genetic algorithm parameters are tested to discuss their impact on prediction accuracy.Results demonstrated that genetic algorithm improved Bi LSTM model is superior to extreme gradient boosting,ARIMA,and artificial neural network,having the highest accuracy of 89%.

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.

Multivariate Two-stage Adaptive-stacking Prediction of Regional Integrated Energy System

To reduce environmental pollution and improve the efficiency of cascaded energy utilization, regional integrated energy system(RIES) has received extensive attention. An accurate multi-energy load prediction is significant for RIES as it enables stakeholders to make effective decisions for carbon peaking and carbon neutrality goals. To this end, this paper proposes a multivariate two-stage adaptive-stacking prediction(M2ASP) framework. First, a preprocessing module based on ensemble learning is proposed. The input data are preprocessed to provide a reliable database for M2ASP, and highly correlated input variables of multi-energy load prediction are determined. Then, the load prediction results of four predictors are adaptively combined in the first stage of M2ASP to enhance generalization ability. Predictor hyper-parameters and intermediate data sets of M2ASP are trained with a metaheuristic method named collaborative atomic chaotic search(CACS) to achieve the adaptive staking of M2ASP. Finally, a prediction correction of the peak load consumption period is conducted in the second stage of M2ASP. The case studies indicate that the proposed framework has higher prediction accuracy, generalization ability, and stability than other benchmark prediction models.

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.

Association of polymorphisms in C1orf106,IL1RN,and IL10 with post-induction infliximab trough level in Crohn’s disease patients

Background:Trough levels of the post-induction serum infliximab(IFX)are associated with short-term and long-term responses of Crohn’s disease patients to IFX,but the inter-individual differences are large.We aimed to elucidate whether single gene polymorphisms(SNPs)within FCGR3A,ATG16L1,C1orf106,OSM,OSMR,NF-jB1,IL1RN,and IL10 partially account for these differences and employed a multivariate regression model to predict patients’post-induction IFX levels.Methods:The retrospective study included 189 Crohn’s disease patients undergoing IFX therapy.Post-induction IFX levels were measured and 41 tag SNPs within eight genes were genotyped.Associations between SNPs and IFX levels were analysed.Then,a multivariate logistic-regression model was developed to predict whether the patients’IFX levels achieved the threshold of therapy(3 lg/mL).Results:Six SNPs(rs7587051,rs143063741,rs442905,rs59457695,rs3213448,and rs3021094)were significantly associated with the post-induction IFX trough level(P=0.015,P<0.001,P=0.046,P=0.022,P=0.011,P=0.013,respectively).A multivariate prediction model of the IFX level was established by baseline albumin(P=0.002),rs442905(P=0.025),rs59457695(P=0.049),rs3213448(P=0.056),and rs3021094(P=0.047).The area under the receiver operating characteristic curve(AUROC)of this prediction model in a representative training dataset was 0.758.This result was verified in a representative testing dataset,with an AUROC of 0.733.Conclusions:Polymorphisms in C1orf106,IL1RN,and IL10 play an important role in the variability of IFX post-induction levels,as indicated in this multivariate prediction model of IFX levels with fair performance.

Sequential Big Data-Based Macroeconomic Forecast for Industrial Value Added

Macroeconomic situation is the overall performance of a country’s and regional economic situation.At present,the vast majority of macroeconomic indicators are obtained through sampling surveys,step-by-step reporting,statistical calculations,and other processes,which are publicly released by the Statistical Bureau.There are some shortcomings,such as lag and non-authenticity.Timely forecasting and early warning of macroeconomic trends are the important needs of government affairs.However,the timeliness of data has a direct impact on government decision-making.In this paper,the high frequency and relatively accurate big data sources are adopted to construct a multivariate regression prediction model for traditional national economic accounting indicators(such as industrial value added above the scale of Hefei),which is different from the traditional time series prediction model such as ARIMA model.Based on the macroeconomic prediction model of time series big data,multi-latitude data sources,sequential update,verification set screening model and other strategies are used to provide more reliable,timely,and easy-to-understand forecasting values of national economic accounting indicators.At the same time,the potential influencing factors of macroeconomic indicators are excavated to provide data and theoretical basis for macroeconomic analysis and decision-making.

The predictive value of monocyte to HDL cholesterol ratio on clinical outcomes in type 2 diabetes mellitus patients undergoing elective percutaneous coronary intervention

Background Monocyte to high density lipoprotein ratio（MHR） has been considered as a novel parameter related with adverse renal and cardiovascular outcomes.In this study we investigated the association of MHR with major adverse clinical events（MACEs） in patients with type 2 diabetes mellitus（T2DM） undergoing elective percutaneous coronary intervention（PCI）.Methods Consecutive T2 DM patients treated with elective PCI were prospectively recruited between July 2008-January 2016 in Department of Cardiology of Panyu Central Hospital.Subjects were categorized into two groups：as patients who developed MACEs（MACEs＋） and patients who did not develop MACEs（MACEs-） during hospitalization.MACEs were defined as the composite end points,including all-cause mortality,or acute heart failure,or target vessel revascularization,or stroke or recurrent angina.Results A total of 418 patients were included in the study.64 patients developed MACEs（15.3%）.In the MACEs（＋） patients,monocytes were higher（1.12 [0.78-1.42] vs.0.72 [0.68-0.92] 109/L,P 〈 0.01） and HDL cholesterol levels were lower（0.87 [0.72-1.21] vs.0.96 [0.81-1.11] mmol/L,P 〈 0.01）.In addition,MHR was significantly higher in the MACEs（＋） group（1.12 [0.91-2.09] vs.0.73[0.54-0.93] 109 mmol/L,P 〈 0.01）.The cutoff value of MHR for predicting MACEs was 22,with a sensitivity of 81% and a specificity of 75.1%（area under the curve0.79,P 〈 0.001）.In multivariate logistic regression analysis,MHR remained an independent factor correlated with MACEs（OR = 3.97,95%CI = 1.38-11.5,P 〈 0.01）.Conclusion Higher MHR levels may predict MACEsdevelopment after elective PCI in T2 DM patients.