The rate of the total electron content(TEC)change index(ROTI)can be regarded as an effective indicator of the level of ionospheric scintillation,in particular in low and high latitude regions.An accurate prediction of...The rate of the total electron content(TEC)change index(ROTI)can be regarded as an effective indicator of the level of ionospheric scintillation,in particular in low and high latitude regions.An accurate prediction of the ROTI is essential to reduce the impact of the ionospheric scintillation on earth observation systems,such as the global navigation satellite systems.However,it is difficult to predict the ROTI with high accuracy because of the complexity of the ionosphere.In this study,advanced machine learning methods have been investigated for ROTI prediction over a station at high-latitude in Canada.These methods are used to predict the ROTI in the next 5 minutes using the data derived from the past 15 minutes at the same location.Experimental results show that the method of the bidirectional gated recurrent unit network(BGRU)outperforms the other six approaches tested in the research.It is also confirmed that the RMSEs of the predicted ROTI using the BGRU method in all four seasons of 2017 are less than 0.05 TECU/min.It is demonstrated that the BGRU method exhibits a high level of robustness in dealing with abrupt solar activities.展开更多
Uncertainty quantification(UQ)is rapidly becoming a sine qua non for all forms of computational science out of which actionable outcomes are anticipated.Much of the microscopic world of atoms and molecules has remaine...Uncertainty quantification(UQ)is rapidly becoming a sine qua non for all forms of computational science out of which actionable outcomes are anticipated.Much of the microscopic world of atoms and molecules has remained immune to these developments but due to the fundamental problems of reproducibility and reliability,it is essential that practitioners pay attention to the issues concerned.Here aUQstudy is undertaken of classical molecular dynamics with a particular focus on uncertainties in the high-dimensional force-field parameters,which affect key quantities of interest,including material properties and binding free energy predictions in drug discovery and personalized medicine.Using scalable UQ methods based on active subspaces that invoke machine learning and Gaussian processes,the sensitivity of the input parameters is ranked.Our analyses reveal that the prediction uncertainty is dominated by a small number of the hundreds of interaction potential parameters within the force fields employed.This ranking highlights what forms of interaction control the prediction uncertainty and enables systematic improvements to be made in future optimizations of such parameters.展开更多
An efficient algorithm is proposed for Bayesian model calibration,which is commonly used to estimate the model parameters of non-linear,computationally expensive models using measurement data.The approach is based on ...An efficient algorithm is proposed for Bayesian model calibration,which is commonly used to estimate the model parameters of non-linear,computationally expensive models using measurement data.The approach is based on Bayesian statistics:using a prior distribution and a likelihood,the posterior distribution is obtained through application of Bayes’law.Our novel algorithm to accurately determine this posterior requires significantly fewer discrete model evaluations than traditional Monte Carlo methods.The key idea is to replace the expensive model by an interpolating surrogate model and to construct the interpolating nodal set maximizing the accuracy of the posterior.To determine such a nodal set an extension to weighted Leja nodes is introduced,based on a new weighting function.We prove that the convergence of the posterior has the same rate as the convergence of the model.If the convergence of the posterior is measured in the Kullback–Leibler divergence,the rate doubles.The algorithm and its theoretical properties are verified in three different test cases:analytical cases that confirm the correctness of the theoretical findings,Burgers’equation to show its applicability in implicit problems,and finally the calibration of the closure parameters of a turbulence model to show the effectiveness for computa-tionally expensive problems.展开更多
基金National Key Research Program of China(No.2017YFE0131400)National Natural Science Foundation of China(Nos.41674043,41704038,41874040)+2 种基金Beijing Nova Program(No.xx2017042)Beijing Talents Foundation(No.2017000021223ZK13)CUMT Independent Innovation Project of“Double-First Class”Construction(No.2018ZZ08)。
文摘The rate of the total electron content(TEC)change index(ROTI)can be regarded as an effective indicator of the level of ionospheric scintillation,in particular in low and high latitude regions.An accurate prediction of the ROTI is essential to reduce the impact of the ionospheric scintillation on earth observation systems,such as the global navigation satellite systems.However,it is difficult to predict the ROTI with high accuracy because of the complexity of the ionosphere.In this study,advanced machine learning methods have been investigated for ROTI prediction over a station at high-latitude in Canada.These methods are used to predict the ROTI in the next 5 minutes using the data derived from the past 15 minutes at the same location.Experimental results show that the method of the bidirectional gated recurrent unit network(BGRU)outperforms the other six approaches tested in the research.It is also confirmed that the RMSEs of the predicted ROTI using the BGRU method in all four seasons of 2017 are less than 0.05 TECU/min.It is demonstrated that the BGRU method exhibits a high level of robustness in dealing with abrupt solar activities.
基金funding support from(i)the UK EPSRC for the UK High-End Computing Consortium(EP/R029598/1)the Software Environment for Actionable&VVUQ-evaluated Exascale Applications(SEAVEA)grant(EP/W007762/1)+5 种基金the UK Consortium on Mesoscale Engineering Sciences(UKCOMES grant no.EP/L00030X/1)the Computational Biomedicine at the Exascale(CompBioMedX)grant(EP/X019276/1)(ii)the UK MRC Medical Bioinformatics project(grant no.MR/L016311/1)(iii)the European Commission for EU H2020 CompBioMed2 Center of Excellence(grant no.823712)EU H2020 EXDCI-2 project(grant no.800957)We made use of SuperMUC-NG at Leibniz Supercomputing Center under project COVID-19-SNG1,and the ARCHER2 UK National Supercomputing Service under the SEAVEA grant(EP/W007762/1).
文摘Uncertainty quantification(UQ)is rapidly becoming a sine qua non for all forms of computational science out of which actionable outcomes are anticipated.Much of the microscopic world of atoms and molecules has remained immune to these developments but due to the fundamental problems of reproducibility and reliability,it is essential that practitioners pay attention to the issues concerned.Here aUQstudy is undertaken of classical molecular dynamics with a particular focus on uncertainties in the high-dimensional force-field parameters,which affect key quantities of interest,including material properties and binding free energy predictions in drug discovery and personalized medicine.Using scalable UQ methods based on active subspaces that invoke machine learning and Gaussian processes,the sensitivity of the input parameters is ranked.Our analyses reveal that the prediction uncertainty is dominated by a small number of the hundreds of interaction potential parameters within the force fields employed.This ranking highlights what forms of interaction control the prediction uncertainty and enables systematic improvements to be made in future optimizations of such parameters.
文摘An efficient algorithm is proposed for Bayesian model calibration,which is commonly used to estimate the model parameters of non-linear,computationally expensive models using measurement data.The approach is based on Bayesian statistics:using a prior distribution and a likelihood,the posterior distribution is obtained through application of Bayes’law.Our novel algorithm to accurately determine this posterior requires significantly fewer discrete model evaluations than traditional Monte Carlo methods.The key idea is to replace the expensive model by an interpolating surrogate model and to construct the interpolating nodal set maximizing the accuracy of the posterior.To determine such a nodal set an extension to weighted Leja nodes is introduced,based on a new weighting function.We prove that the convergence of the posterior has the same rate as the convergence of the model.If the convergence of the posterior is measured in the Kullback–Leibler divergence,the rate doubles.The algorithm and its theoretical properties are verified in three different test cases:analytical cases that confirm the correctness of the theoretical findings,Burgers’equation to show its applicability in implicit problems,and finally the calibration of the closure parameters of a turbulence model to show the effectiveness for computa-tionally expensive problems.
