Despite the maturity of ensemble numerical weather prediction(NWP),the resulting forecasts are still,more often than not,under-dispersed.As such,forecast calibration tools have become popular.Among those tools,quantil...Despite the maturity of ensemble numerical weather prediction(NWP),the resulting forecasts are still,more often than not,under-dispersed.As such,forecast calibration tools have become popular.Among those tools,quantile regression(QR)is highly competitive in terms of both flexibility and predictive performance.Nevertheless,a long-standing problem of QR is quantile crossing,which greatly limits the interpretability of QR-calibrated forecasts.On this point,this study proposes a non-crossing quantile regression neural network(NCQRNN),for calibrating ensemble NWP forecasts into a set of reliable quantile forecasts without crossing.The overarching design principle of NCQRNN is to add on top of the conventional QRNN structure another hidden layer,which imposes a non-decreasing mapping between the combined output from nodes of the last hidden layer to the nodes of the output layer,through a triangular weight matrix with positive entries.The empirical part of the work considers a solar irradiance case study,in which four years of ensemble irradiance forecasts at seven locations,issued by the European Centre for Medium-Range Weather Forecasts,are calibrated via NCQRNN,as well as via an eclectic mix of benchmarking models,ranging from the naïve climatology to the state-of-the-art deep-learning and other non-crossing models.Formal and stringent forecast verification suggests that the forecasts post-processed via NCQRNN attain the maximum sharpness subject to calibration,amongst all competitors.Furthermore,the proposed conception to resolve quantile crossing is remarkably simple yet general,and thus has broad applicability as it can be integrated with many shallow-and deep-learning-based neural networks.展开更多
In this study,mung bean liquid(MBL)was prepared by wet milling and then naturally fermented for 0,24,48,72,and 96 h.Afterward,the starch extracted from MBL was analyzed.The pH of MBL decreases from 6.54 to 4.91 during...In this study,mung bean liquid(MBL)was prepared by wet milling and then naturally fermented for 0,24,48,72,and 96 h.Afterward,the starch extracted from MBL was analyzed.The pH of MBL decreases from 6.54 to 4.91 during fermentation.Moreover,the mung bean starch(MBS)granules changed significantly with pitted surfaces,pores and channels.The particle size,molecular weight(Mw)and long chain amylopectin(B2)content of the MBS decreased,whereas the short chain(A)increased from 20.57% to 22.70% with increasing fermentation time.Natural fermentation had no effect on the crystalline type(A)of the MBS but increased the relative crystallinity from 26.53% to 29.23%.However,the intensity ratios of the 995/1022 cm^(-1) and 1047/1022 cm^(-1) FTIR spectra increased with fermentation time,reaching maxima of 0.95 and 0.80 after 48 h,respectively.In addition,Fermented MBS increased the trough viscosity(TV),final viscosity(FV),elastic modulus(G′)and viscous modulus(G″)but decreased the peak viscosity(PV)and breakdown(BD).Thermal properties revealed that fermentation increased the gelatinization enthalpy(ΔH)from 9.49 to 10.26 J/g.展开更多
基金supported by the National Natural Science Foundation of China (Project No.42375192)the China Meteorological Administration Climate Change Special Program (CMA-CCSP+1 种基金Project No.QBZ202315)support by the Vector Stiftung through the Young Investigator Group"Artificial Intelligence for Probabilistic Weather Forecasting."
文摘Despite the maturity of ensemble numerical weather prediction(NWP),the resulting forecasts are still,more often than not,under-dispersed.As such,forecast calibration tools have become popular.Among those tools,quantile regression(QR)is highly competitive in terms of both flexibility and predictive performance.Nevertheless,a long-standing problem of QR is quantile crossing,which greatly limits the interpretability of QR-calibrated forecasts.On this point,this study proposes a non-crossing quantile regression neural network(NCQRNN),for calibrating ensemble NWP forecasts into a set of reliable quantile forecasts without crossing.The overarching design principle of NCQRNN is to add on top of the conventional QRNN structure another hidden layer,which imposes a non-decreasing mapping between the combined output from nodes of the last hidden layer to the nodes of the output layer,through a triangular weight matrix with positive entries.The empirical part of the work considers a solar irradiance case study,in which four years of ensemble irradiance forecasts at seven locations,issued by the European Centre for Medium-Range Weather Forecasts,are calibrated via NCQRNN,as well as via an eclectic mix of benchmarking models,ranging from the naïve climatology to the state-of-the-art deep-learning and other non-crossing models.Formal and stringent forecast verification suggests that the forecasts post-processed via NCQRNN attain the maximum sharpness subject to calibration,amongst all competitors.Furthermore,the proposed conception to resolve quantile crossing is remarkably simple yet general,and thus has broad applicability as it can be integrated with many shallow-and deep-learning-based neural networks.
基金financially supported by the Regional Innovation and Development Joint Fund of National Natural Science Foundation of China(U22A20537)the Key Research and Development Project of Henan province(231111113200)the University Scientific and Tech-nological Innovation Teams in Henan province(23IRTSTHN029).
文摘In this study,mung bean liquid(MBL)was prepared by wet milling and then naturally fermented for 0,24,48,72,and 96 h.Afterward,the starch extracted from MBL was analyzed.The pH of MBL decreases from 6.54 to 4.91 during fermentation.Moreover,the mung bean starch(MBS)granules changed significantly with pitted surfaces,pores and channels.The particle size,molecular weight(Mw)and long chain amylopectin(B2)content of the MBS decreased,whereas the short chain(A)increased from 20.57% to 22.70% with increasing fermentation time.Natural fermentation had no effect on the crystalline type(A)of the MBS but increased the relative crystallinity from 26.53% to 29.23%.However,the intensity ratios of the 995/1022 cm^(-1) and 1047/1022 cm^(-1) FTIR spectra increased with fermentation time,reaching maxima of 0.95 and 0.80 after 48 h,respectively.In addition,Fermented MBS increased the trough viscosity(TV),final viscosity(FV),elastic modulus(G′)and viscous modulus(G″)but decreased the peak viscosity(PV)and breakdown(BD).Thermal properties revealed that fermentation increased the gelatinization enthalpy(ΔH)from 9.49 to 10.26 J/g.
