To realize data sharing,and to fully use the data value,breaking the data island between institutions to realize data collaboration has become a new sharing mode.This paper proposed a distributed data security sharing...To realize data sharing,and to fully use the data value,breaking the data island between institutions to realize data collaboration has become a new sharing mode.This paper proposed a distributed data security sharing scheme based on C/S communication mode,and constructed a federated learning architecture that uses differential privacy technology to protect training parameters.Clients do not need to share local data,and they only need to upload the trained model parameters to achieve data sharing.In the process of training,a distributed parameter update mechanism is introduced.The server is mainly responsible for issuing training commands and parameters,and aggregating the local model parameters uploaded by the clients.The client mainly uses the stochastic gradient descent algorithm for gradient trimming,updates,and transmits the trained model parameters back to the server after differential processing.To test the performance of the scheme,in the application scenario where many medical institutions jointly train the disease detection system,the model is tested from multiple perspectives by taking medical data as an example.From the testing results,we can know that for this specific test dataset,when the parameters are properly configured,the lowest prediction accuracy rate is 90.261%and the highest accuracy rate is up to 94.352.It shows that the performance of the model is good.The results also show that this scheme realizes data sharing while protecting data privacy,completes accurate prediction of diseases,and has a good effect.展开更多
Mount Fouimba and Mount Goma (Seguela) greenstone belts petro-structural studies combine remote sensing, geophysics, petrography and structural analysis. In view of establishing mapping details of paleoproterozoic geo...Mount Fouimba and Mount Goma (Seguela) greenstone belts petro-structural studies combine remote sensing, geophysics, petrography and structural analysis. In view of establishing mapping details of paleoproterozoic geological formations, geological setting rocks observed are essentially magmatic formations, such as two-mica granite, granodiorites, and porphyritic basalts;and a few metamorphics which are metatonalite, amphibolites and amphibo-lopyroxenites. Remote sensing, such as Landsat 8 OLI satellite imagery and geophysical data, has been combined to show regional NNE-SW shear zone. Tectonic structures and microstructures have enabled to identify two main deformation phases: D1 phase corresponding to compression, and D2 is a transpression phase. Mechanisms responsible for deformations are respectively flattening and transpression. Geological formations derived from mantle origin but contain crustal components, and their tectonic setting occurred during subduction.展开更多
An experimental advanced superconducting tokamak （EAST） operation window with the lower hybrid current drive （LHCD） in H-mode is estimated by using a eore-SOL-divertor （C-S-D） model validated by the present EAST...An experimental advanced superconducting tokamak （EAST） operation window with the lower hybrid current drive （LHCD） in H-mode is estimated by using a eore-SOL-divertor （C-S-D） model validated by the present EAST divertor experiments. The operation window consists of four limits including two usual limits, one of which is the maximum allowable heat load onto the divertor plate, and two additional limits associated with the LHCD. The predictive EAST operation window is not qualified to fulfill its mission for high input power. To extend the operation window, gas puffing and impurity seeding are presented as two effective methods. In addition, the effect of the LHCD current on the operation window is also discussed. Our numerical analysis results provide a reference for the safe operation of EAST experiments with LHCD in future.展开更多
基金This work was supported by Funding of the Nanjing Institute of Technology(No.KE21-451).
文摘To realize data sharing,and to fully use the data value,breaking the data island between institutions to realize data collaboration has become a new sharing mode.This paper proposed a distributed data security sharing scheme based on C/S communication mode,and constructed a federated learning architecture that uses differential privacy technology to protect training parameters.Clients do not need to share local data,and they only need to upload the trained model parameters to achieve data sharing.In the process of training,a distributed parameter update mechanism is introduced.The server is mainly responsible for issuing training commands and parameters,and aggregating the local model parameters uploaded by the clients.The client mainly uses the stochastic gradient descent algorithm for gradient trimming,updates,and transmits the trained model parameters back to the server after differential processing.To test the performance of the scheme,in the application scenario where many medical institutions jointly train the disease detection system,the model is tested from multiple perspectives by taking medical data as an example.From the testing results,we can know that for this specific test dataset,when the parameters are properly configured,the lowest prediction accuracy rate is 90.261%and the highest accuracy rate is up to 94.352.It shows that the performance of the model is good.The results also show that this scheme realizes data sharing while protecting data privacy,completes accurate prediction of diseases,and has a good effect.
文摘Mount Fouimba and Mount Goma (Seguela) greenstone belts petro-structural studies combine remote sensing, geophysics, petrography and structural analysis. In view of establishing mapping details of paleoproterozoic geological formations, geological setting rocks observed are essentially magmatic formations, such as two-mica granite, granodiorites, and porphyritic basalts;and a few metamorphics which are metatonalite, amphibolites and amphibo-lopyroxenites. Remote sensing, such as Landsat 8 OLI satellite imagery and geophysical data, has been combined to show regional NNE-SW shear zone. Tectonic structures and microstructures have enabled to identify two main deformation phases: D1 phase corresponding to compression, and D2 is a transpression phase. Mechanisms responsible for deformations are respectively flattening and transpression. Geological formations derived from mantle origin but contain crustal components, and their tectonic setting occurred during subduction.
基金supported by National Natural Science Foundation of China(Nos.11105176 and 11105224)
文摘An experimental advanced superconducting tokamak （EAST） operation window with the lower hybrid current drive （LHCD） in H-mode is estimated by using a eore-SOL-divertor （C-S-D） model validated by the present EAST divertor experiments. The operation window consists of four limits including two usual limits, one of which is the maximum allowable heat load onto the divertor plate, and two additional limits associated with the LHCD. The predictive EAST operation window is not qualified to fulfill its mission for high input power. To extend the operation window, gas puffing and impurity seeding are presented as two effective methods. In addition, the effect of the LHCD current on the operation window is also discussed. Our numerical analysis results provide a reference for the safe operation of EAST experiments with LHCD in future.
