In this paper,we give the double inequalities of a special quasi-arithmetic mean E(a,b)and the Seifert-like elliptic integral mean V(a,b)under the combination of generalized Heronian mean Hw(a,b),harmonic mean H(a,b)a...In this paper,we give the double inequalities of a special quasi-arithmetic mean E(a,b)and the Seifert-like elliptic integral mean V(a,b)under the combination of generalized Heronian mean Hw(a,b),harmonic mean H(a,b)and arithmetic mean A(a,b).As applications,we obtain two new optimal bounds for the second kind of complete elliptic integral".Finally,numerical experiments are performed to compare these with some known bounds of",and the improved effect is verified.展开更多
针对目前城市道路场景中行道树提取方法需要设置的参数较多以及树冠点云相互重叠难以精确分割的问题,文章采用一种行道树提取与单株木分割算法。首先通过布料滤波算法从原始点云中移除地面点,并利用半径滤波滤除离群点,去除地面点和噪...针对目前城市道路场景中行道树提取方法需要设置的参数较多以及树冠点云相互重叠难以精确分割的问题,文章采用一种行道树提取与单株木分割算法。首先通过布料滤波算法从原始点云中移除地面点,并利用半径滤波滤除离群点,去除地面点和噪声点对行道树提取的影响;然后通过增加PointNet++网络的点集抽象模块(set abstraction,SA)提高模型特征提取能力,使模型更适用于行道树点云的提取,并利用改进后的网络从原始点云中提取行道树点云;最后结合密度聚类算法(density-based spatial clustering of applications with noise,DBSCAN)与K-Means算法对相互重叠的行道树点云进行分割,得到单株木信息。为验证该方法的有效性,以北京永昌路道路数据集进行训练测试。结果表明:改进后模型的行道树点云平均提取精度和交并比(intersection over union,IoU)分别提高了9.2%和15.1%,达到了94.5%、0.916;单木分割平均精度达到了91.3%。展开更多
In the dispersed computing environment driven by intelligent networks,intrusion detection faces significant challenges.This paper proposes a multilayer decentralized federated learning framework based on mean field ga...In the dispersed computing environment driven by intelligent networks,intrusion detection faces significant challenges.This paper proposes a multilayer decentralized federated learning framework based on mean field game theory(MFG-DFL).The framework organizes networked computing points(NCPs)into a three-layer collaborative architecture,and innovatively introduces MFG theory to model the complex dynamic interactions,which among large-scale NCPs as a game between a representative NCP and the mean field.By solving the coupled HJB and FPK equations,we design a dynamic incentive mechanism to fairly quantify and reward NCP contributions,thus aligning individual rationality with the global objectives of the system.The simulation results on the CICIoT2023 data set demonstrate the outstanding performance of the proposed framework.Specifically,it achieves an intrusion detection accuracy of 81.09%in highly non-IID scenarios,showcasing a well-balanced trade-off between computational efficiency and performance enhancement.展开更多
In Global Navigation Satellite System(GNSS)meteo rology,the atmospheric weighted mean temperatu re(T_(m))is a critical intermediate parameter for converting zenith wet delay(ZWD)to precipitable water vapor(PWV),essent...In Global Navigation Satellite System(GNSS)meteo rology,the atmospheric weighted mean temperatu re(T_(m))is a critical intermediate parameter for converting zenith wet delay(ZWD)to precipitable water vapor(PWV),essential for accurate atmospheric water content estimation.However,global models often overlook regional climatic variability,leading to reduced accuracy in localized applications.This study introduces an improved T_(m)model developed using radiosonde observations across Iran and GNSS radio occultation(RO)profiles from CHAMP,GRACE,MetOp-A/B/C,COSMIC,TerraSAR-X,and TanDEM-X missions collected between 2007 and 2022.A novel integral formulation was proposed to estimate T_(m)more accurately by incorporating vertical water vapor distribution and temperature linearity.Based on this formulation,three regional T_(m)models were constructed using annual,semiannual,and diurnal periodicities,along with surface temperature(T_(s)),each varying in structure and complexity.Validation against independent radiosonde observations from 2022 showed that Models Two and Three outperformed the Bevis model,reducing RMSE by 30.7%.When evaluated against GNSS RO profiles,Model One—excluding T_(s)due to its inaccessibility in RO data—yielded the highest accuracy,with a 42.6%improvement in RMSE over the Bevis model.To evaluate the practical effectiveness of the proposed T_(m)model,PWV was derived from GNSS data at the tehn and tabz stations during the second half of 2022and compared with PWV values obtained from co-located radiosonde observations in Tehran and Tabriz.Using T_(m)from Model One improved PWV estimation compared to the Bevis model,reducing RMSE and MAE by up to 54%and 53.8%in Tabriz and 50.6%and 52.9%in Tehran,respectively.These results demonstrate that regionalized T_(m)modeling,particularly approaches that avoid dependence on T_(s),can significantly enhance GNSS-based PWV estimation in areas with limited surface data.展开更多
Multi-dimensional arrays are referred to as tensors.Tensor-valued predictors are commonly encountered in modern biomedical applications,such as electroencephalogram(EEG),magnetic resonance imaging(MRI),functional MRI(...Multi-dimensional arrays are referred to as tensors.Tensor-valued predictors are commonly encountered in modern biomedical applications,such as electroencephalogram(EEG),magnetic resonance imaging(MRI),functional MRI(fMRI),diffusion-weighted MRI,and longitudinal health data.In survival analysis,it is both important and challenging to integrate clinically relevant information,such as gender,age,and disease state along with medical imaging tensor data or longitudinal health data to predict disease outcomes.Most existing higher-order sufficient dimension reduction regressions for matrix-or array-valued data focus solely on tensor data,often neglecting established clinical covariates that are readily available and known to have predictive value.Based on the idea of Folded-Minimum Average Variance Estimation(Folded-MAVE:Xue and Yin,2014),the authors propose a new method,Partial Dimension Folded-MAVE(PF-MAVE),to address regression mean functions with tensor-valued covariates while simultaneously incorporating clinical covariates,which are typically categorical variables.Theorems and simulation studies demonstrate the importance of incorporating these categorical clinical predictors.A survival analysis of a longitudinal study of primary biliary cirrhosis(PBC)data is included for illustration of the proposed method.展开更多
文摘In this paper,we give the double inequalities of a special quasi-arithmetic mean E(a,b)and the Seifert-like elliptic integral mean V(a,b)under the combination of generalized Heronian mean Hw(a,b),harmonic mean H(a,b)and arithmetic mean A(a,b).As applications,we obtain two new optimal bounds for the second kind of complete elliptic integral".Finally,numerical experiments are performed to compare these with some known bounds of",and the improved effect is verified.
文摘针对目前城市道路场景中行道树提取方法需要设置的参数较多以及树冠点云相互重叠难以精确分割的问题,文章采用一种行道树提取与单株木分割算法。首先通过布料滤波算法从原始点云中移除地面点,并利用半径滤波滤除离群点,去除地面点和噪声点对行道树提取的影响;然后通过增加PointNet++网络的点集抽象模块(set abstraction,SA)提高模型特征提取能力,使模型更适用于行道树点云的提取,并利用改进后的网络从原始点云中提取行道树点云;最后结合密度聚类算法(density-based spatial clustering of applications with noise,DBSCAN)与K-Means算法对相互重叠的行道树点云进行分割,得到单株木信息。为验证该方法的有效性,以北京永昌路道路数据集进行训练测试。结果表明:改进后模型的行道树点云平均提取精度和交并比(intersection over union,IoU)分别提高了9.2%和15.1%,达到了94.5%、0.916;单木分割平均精度达到了91.3%。
基金National Science Foundation Project of China(62436004,62372317)National Key Research and Development Program of China(2023YFC3331702)。
文摘In the dispersed computing environment driven by intelligent networks,intrusion detection faces significant challenges.This paper proposes a multilayer decentralized federated learning framework based on mean field game theory(MFG-DFL).The framework organizes networked computing points(NCPs)into a three-layer collaborative architecture,and innovatively introduces MFG theory to model the complex dynamic interactions,which among large-scale NCPs as a game between a representative NCP and the mean field.By solving the coupled HJB and FPK equations,we design a dynamic incentive mechanism to fairly quantify and reward NCP contributions,thus aligning individual rationality with the global objectives of the system.The simulation results on the CICIoT2023 data set demonstrate the outstanding performance of the proposed framework.Specifically,it achieves an intrusion detection accuracy of 81.09%in highly non-IID scenarios,showcasing a well-balanced trade-off between computational efficiency and performance enhancement.
文摘In Global Navigation Satellite System(GNSS)meteo rology,the atmospheric weighted mean temperatu re(T_(m))is a critical intermediate parameter for converting zenith wet delay(ZWD)to precipitable water vapor(PWV),essential for accurate atmospheric water content estimation.However,global models often overlook regional climatic variability,leading to reduced accuracy in localized applications.This study introduces an improved T_(m)model developed using radiosonde observations across Iran and GNSS radio occultation(RO)profiles from CHAMP,GRACE,MetOp-A/B/C,COSMIC,TerraSAR-X,and TanDEM-X missions collected between 2007 and 2022.A novel integral formulation was proposed to estimate T_(m)more accurately by incorporating vertical water vapor distribution and temperature linearity.Based on this formulation,three regional T_(m)models were constructed using annual,semiannual,and diurnal periodicities,along with surface temperature(T_(s)),each varying in structure and complexity.Validation against independent radiosonde observations from 2022 showed that Models Two and Three outperformed the Bevis model,reducing RMSE by 30.7%.When evaluated against GNSS RO profiles,Model One—excluding T_(s)due to its inaccessibility in RO data—yielded the highest accuracy,with a 42.6%improvement in RMSE over the Bevis model.To evaluate the practical effectiveness of the proposed T_(m)model,PWV was derived from GNSS data at the tehn and tabz stations during the second half of 2022and compared with PWV values obtained from co-located radiosonde observations in Tehran and Tabriz.Using T_(m)from Model One improved PWV estimation compared to the Bevis model,reducing RMSE and MAE by up to 54%and 53.8%in Tabriz and 50.6%and 52.9%in Tehran,respectively.These results demonstrate that regionalized T_(m)modeling,particularly approaches that avoid dependence on T_(s),can significantly enhance GNSS-based PWV estimation in areas with limited surface data.
文摘Multi-dimensional arrays are referred to as tensors.Tensor-valued predictors are commonly encountered in modern biomedical applications,such as electroencephalogram(EEG),magnetic resonance imaging(MRI),functional MRI(fMRI),diffusion-weighted MRI,and longitudinal health data.In survival analysis,it is both important and challenging to integrate clinically relevant information,such as gender,age,and disease state along with medical imaging tensor data or longitudinal health data to predict disease outcomes.Most existing higher-order sufficient dimension reduction regressions for matrix-or array-valued data focus solely on tensor data,often neglecting established clinical covariates that are readily available and known to have predictive value.Based on the idea of Folded-Minimum Average Variance Estimation(Folded-MAVE:Xue and Yin,2014),the authors propose a new method,Partial Dimension Folded-MAVE(PF-MAVE),to address regression mean functions with tensor-valued covariates while simultaneously incorporating clinical covariates,which are typically categorical variables.Theorems and simulation studies demonstrate the importance of incorporating these categorical clinical predictors.A survival analysis of a longitudinal study of primary biliary cirrhosis(PBC)data is included for illustration of the proposed method.
