Artificial Intelligence(AI)in healthcare enables predicting diabetes using data-driven methods instead of the traditional ways of screening the disease,which include hemoglobin A1c(HbA1c),oral glucose tolerance test(O...Artificial Intelligence(AI)in healthcare enables predicting diabetes using data-driven methods instead of the traditional ways of screening the disease,which include hemoglobin A1c(HbA1c),oral glucose tolerance test(OGTT),and fasting plasma glucose(FPG)screening techniques,which are invasive and limited in scale.Machine learning(ML)and deep neural network(DNN)models that use large datasets to learn the complex,nonlinear feature interactions,but the conventional ML algorithms are data sensitive and often show unstable predictive accuracy.Conversely,DNN models are more robust,though the ability to reach a high accuracy rate consistently on heterogeneous datasets is still an open challenge.For predicting diabetes,this work proposed a hybrid DNN approach by integrating a bidirectional long short-term memory(BiLSTM)network with a bidirectional gated recurrent unit(BiGRU).A robust DL model,developed by combining various datasets with weighted coefficients,dense operations in the connection of deep layers,and the output aggregation using batch normalization and dropout functions to avoid overfitting.The goal of this hybrid model is better generalization and consistency among various datasets,which facilitates the effective management and early intervention.The proposed DNN model exhibits an excellent predictive performance as compared to the state-of-the-art and baseline ML and DNN models for diabetes prediction tasks.The robust performance indicates the possible usefulness of DL-based models in the development of disease prediction in healthcare and other areas that demand high-quality analytics.展开更多
As an important resource in data link,time slots should be strategically allocated to enhance transmission efficiency and resist eavesdropping,especially considering the tremendous increase in the number of nodes and ...As an important resource in data link,time slots should be strategically allocated to enhance transmission efficiency and resist eavesdropping,especially considering the tremendous increase in the number of nodes and diverse communication needs.It is crucial to design control sequences with robust randomness and conflict-freeness to properly address differentiated access control in data link.In this paper,we propose a hierarchical access control scheme based on control sequences to achieve high utilization of time slots and differentiated access control.A theoretical bound of the hierarchical control sequence set is derived to characterize the constraints on the parameters of the sequence set.Moreover,two classes of optimal hierarchical control sequence sets satisfying the theoretical bound are constructed,both of which enable the scheme to achieve maximum utilization of time slots.Compared with the fixed time slot allocation scheme,our scheme reduces the symbol error rate by up to 9%,which indicates a significant improvement in anti-interference and eavesdropping capabilities.展开更多
A three-dimensional(3D)electromagnetic(EM)inversion algorithm based on the nonlinear conjugate gradient(NLCG)method and a two-color plane Gauss-Seidel(GS)multigrid(MG)forward solver is developed to improve inversion e...A three-dimensional(3D)electromagnetic(EM)inversion algorithm based on the nonlinear conjugate gradient(NLCG)method and a two-color plane Gauss-Seidel(GS)multigrid(MG)forward solver is developed to improve inversion efficiency.The results indicate that the computational efficiency of each inversion can be improved by approximately a factor of three by using the proposed MG solver.First,the accuracy of the MG solver is validated through a test on a synthetic model.Next,the numerical performance of the inversion algorithm is evaluated using this model.Finally,the inversion algorithm is applied to a field EM data collected at the Beiya gold polymetallic ore district.A 3D resistivity model is obtained,and the formation process of the metal ore is analyzed.展开更多
With the evolution of next-generation network technologies,the complexity of network management has significantly increased,and the means of network attacks are diversified,bringing new challenges to network traffic c...With the evolution of next-generation network technologies,the complexity of network management has significantly increased,and the means of network attacks are diversified,bringing new challenges to network traffic classification.This paper presents a general AIdriven network traffic classification workflow and elaborates on a traffic data and feature engineering framework.Most importantly,it analyzes the concept and causes of data distribution shifts in ne twork traffic,proposing detection methods and countermeasures.Experimental results on real traffic collected at different time intervals show that application evolution can induce data distribution shifts,which in turn lead to a noticeable degradation in traffic classification performance.Comparative drift detection experiments further confirm that such shifts are more evident over long-term intervals,while short-term traffic remains relatively stable.These findings demonstrate the necessity of incorporating drift-aware mechanisms into AI-driven network traffic classification systems.展开更多
Indeed,a range of systems in the environment requires timely,spatially explicit,and credible information to support its environmental decision-making,but no one observing system can give the complete and reliable meas...Indeed,a range of systems in the environment requires timely,spatially explicit,and credible information to support its environmental decision-making,but no one observing system can give the complete and reliable measures of the Earth system across scales.This review summarizes how the realization of the Compute the Planet is underway in the form of machine learning,remote sensing,and sensor data fusion to generate decision-ready environmental insights.We use the application-first approach,which considers remote sensing,in situ and Internet of Things(IoT)sensing,and physics-based models as complementary streams of evidence with similar strengths and failures.We look critically at how an integrated system can convert heterogeneous observations to action products across three high impact application areas:atmosphere and air quality,water–land–ecosystem dynamics,and hazards.Rapid-response situational awareness,ecosystem condition metrics,drought and flood indicators,exposure maps,and hazard/extreme indicators are key products.The integrated systems to environment interface in three high impact application areas:atmosphere and air quality,water-land-ecosystem dynamics,and hazard Examine Our operational requirements can often determine real-life value such as latency,time stability,smooth degradation in the presence of missing or degraded inputs,and calibrated uncertainty usable in thresholdbased decisions.These pitfalls are common across fields:mismatch in the scale between a point sensor and a gridded product,objectives on proxies in remotely sensed measurements,domain shift in the extremes and changing baselines,and evaluation aspects,which overestimate generalization because of spatiotemporal autocorrelation.Based on these lessons,we present cross-domain proposals for strong validation,uncertainty quantification,provenance,and versioning,as well as fair performance evaluation.We conclude that the next era of environmental intelligence will see a reduction in average accuracy improvement and an increase in terms of robustness,transparency,and operational responsibility,thus allowing the integrated environmental intelligence system to be deployed,which may be relied on to monitor human health,resource allocation,and survival in a more climate-adapted world.展开更多
Security and access control for data storage in 5G industrial Internet collaborative systems are facing significant challenges.The characteristics of 5 G networks,such as low latency and high speed,facilitate data tra...Security and access control for data storage in 5G industrial Internet collaborative systems are facing significant challenges.The characteristics of 5 G networks,such as low latency and high speed,facilitate data transmission in the industrial Internet but also increase vulnerability to attacks like theft and tampering.Moreover,in 5G industrial Internet collaborative system environments,data flows across multiple entities and links,which necessitates a flexible access control model to meet specific data access requirements.Traditional role-based and attribute-based access control mechanisms are difficult to apply in such dynamic application scenarios.To address these challenges,we propose a novel data storage solution for 5G industrial Internet collaborative systems.Similar to existing approaches,it provides integrity and confidentiality protection for transmitted data.In terms of security,only authenticated data owners and users can obtain file decryption keys,preventing malicious attackers from data forgery.Regarding access control,decryption is permitted only to authorized data users,safeguarding against unauthorized file access.Furthermore,by introducing an attribute-based encryption mechanism,only data users with specific attributes can decrypt files.In terms of efficiency,our approach utilizes bilinear and modular exponentiation operations solely during the authentication process.For handling substantial data loads,lightweight cryptographic algorithms are employed.Consequently,our solution achieves higher efficiency compared with other known methods.Experimental results demonstrate the feasibility of our approach in real-world applications.展开更多
A dual‑task parallel machine learning framework was developed by integrating a convolutional autoencoder(CAE)and a fully connected neural network(FCNN)via the gradient‑coupled mechanism,enabling simultaneous data comp...A dual‑task parallel machine learning framework was developed by integrating a convolutional autoencoder(CAE)and a fully connected neural network(FCNN)via the gradient‑coupled mechanism,enabling simultaneous data compression‑reconstruction and structural damage identification.Under the condition where 40% of the sensor nodes are missing,the model successfully reconstructs the full sensor network with an R^(2) of 0.916 and normalized root mean square error(NRMSE)of 0.0288.Even under significant noise contamination with an SNR of 12 dB,the model maintains strong reconstruction performance,achieving a R^(2) of 0.910 and NRMSE of 0.0253.Forty‑six structural damage scenarios were simulated using the scaled bridge model.The accuracy of spatial localization and quantification of the damage severity using the framework exceeds 99.3%.The proposed framework reduces the training time by 54.4%and iteration counts by 45.5% compared to conventional two‑stage machine learning approaches,demonstrating the efficiency of gradient‑coupled optimization.展开更多
Parkinson’s disease(PD)is a debilitating neurological disorder affecting over 10 million people worldwide.PD classification models using voice signals as input are common in the literature.It is believed that using d...Parkinson’s disease(PD)is a debilitating neurological disorder affecting over 10 million people worldwide.PD classification models using voice signals as input are common in the literature.It is believed that using deep learning algorithms further enhances performance;nevertheless,it is challenging due to the nature of small-scale and imbalanced PD datasets.This paper proposed a convolutional neural network-based deep support vector machine(CNN-DSVM)to automate the feature extraction process using CNN and extend the conventional SVM to a DSVM for better classification performance in small-scale PD datasets.A customized kernel function reduces the impact of biased classification towards the majority class(healthy candidates in our consideration).An improved generative adversarial network(IGAN)was designed to generate additional training data to enhance the model’s performance.For performance evaluation,the proposed algorithm achieves a sensitivity of 97.6%and a specificity of 97.3%.The performance comparison is evaluated from five perspectives,including comparisons with different data generation algorithms,feature extraction techniques,kernel functions,and existing works.Results reveal the effectiveness of the IGAN algorithm,which improves the sensitivity and specificity by 4.05%–4.72%and 4.96%–5.86%,respectively;and the effectiveness of the CNN-DSVM algorithm,which improves the sensitivity by 1.24%–57.4%and specificity by 1.04%–163%and reduces biased detection towards the majority class.The ablation experiments confirm the effectiveness of individual components.Two future research directions have also been suggested.展开更多
Understanding the dynamic traffic and usage characteristics of data services in cellular networks is important for optimising network resources and improving user experience.Recent studies have illustrated traffic cha...Understanding the dynamic traffic and usage characteristics of data services in cellular networks is important for optimising network resources and improving user experience.Recent studies have illustrated traffic characteristics from specific perspectives,such as user behaviour,device type,and applications.In this paper,we present the results of our study from a different perspective,namely service providers,to reveal the traffic characteristics of cellular data networks.Our study is based on traffic data collected over a five-day period from a leading mobile operator's core network in China.We propose a Zipf-like model to characterise the distributions of the traffic volume,subscribers,and requests among service providers.Nine distinct diurnal traffic patterns of service providers are identified by formulating and solving a time series clustering problem.Our work differs from previous related works in that we perform measurements on a large quantity of data covering 2.2 billion traffic records,and we first explore the traffic patterns of thousands of service providers.Results of our study present mobile Internet participants with a better understanding of the traffic and usage characteristics of service providers,which play a critical role in the mobile Internet era.展开更多
In order to address the widespread data shortage problem in battery research,this paper proposes a generative adversarial network model that combines it with deep convolutional networks,the Wasserstein distance,and th...In order to address the widespread data shortage problem in battery research,this paper proposes a generative adversarial network model that combines it with deep convolutional networks,the Wasserstein distance,and the gradient penalty to achieve data augmentation.To lower the threshold for implementing the proposed method,transfer learning is further introduced.The W-DC-GAN-GP-TL framework is thereby formed.This framework is evaluated on 3 different publicly available datasets to judge the quality of generated data.Through visual comparisons and the examination of two visualization methods(probability density function(PDF)and principal component analysis(PCA)),it is demonstrated that the generated data is hard to distinguish from the real data.The application of generated data for training a battery state model using transfer learning is further evaluated.Specifically,Bi-GRU-based and Transformer-based methods are implemented on 2 separate datasets for estimating state of health(SOH)and state of charge(SOC),respectively.The results indicate that the proposed framework demonstrates satisfactory performance in different scenarios:for the data replacement scenario,where real data are removed and replaced with generated data,the state estimator accuracy decreases only slightly;for the data enhancement scenario,the estimator accuracy is further improved.The estimation accuracy of SOH and SOC is as low as 0.69%and 0.58%root mean square error(RMSE)after applying the proposed framework.This framework provides a reliable method for enriching battery measurement data.It is a generalized framework capable of generating a variety of time series data.展开更多
Earth’s internal core and crustal magnetic fields,as measured by geomagnetic satellites like MSS-1(Macao Science Satellite-1)and Swarm,are vital for understanding core dynamics and tectonic evolution.To model these i...Earth’s internal core and crustal magnetic fields,as measured by geomagnetic satellites like MSS-1(Macao Science Satellite-1)and Swarm,are vital for understanding core dynamics and tectonic evolution.To model these internal magnetic fields accurately,data selection based on specific criteria is often employed to minimize the influence of rapidly changing current systems in the ionosphere and magnetosphere.However,the quantitative impact of various data selection criteria on internal geomagnetic field modeling is not well understood.This study aims to address this issue and provide a reference for constructing and applying geomagnetic field models.First,we collect the latest MSS-1 and Swarm satellite magnetic data and summarize widely used data selection criteria in geomagnetic field modeling.Second,we briefly describe the method to co-estimate the core,crustal,and large-scale magnetospheric fields using satellite magnetic data.Finally,we conduct a series of field modeling experiments with different data selection criteria to quantitatively estimate their influence.Our numerical experiments confirm that without selecting data from dark regions and geomagnetically quiet times,the resulting internal field differences at the Earth’s surface can range from tens to hundreds of nanotesla(nT).Additionally,we find that the uncertainties introduced into field models by different data selection criteria are significantly larger than the measurement accuracy of modern geomagnetic satellites.These uncertainties should be considered when utilizing constructed magnetic field models for scientific research and applications.展开更多
As the world’s largest digital economy,China has a significant demand for data centers,which are energy-intensive.With an annual growth rate of 28%in installed capacity,these centers are primarily located in the deve...As the world’s largest digital economy,China has a significant demand for data centers,which are energy-intensive.With an annual growth rate of 28%in installed capacity,these centers are primarily located in the developed eastern region,where land and energy resources are limited.This localization poses a major challenge to the industry’s net-zero goal.To address this,China has launched a bold initiative to relocate data centers to the western region,leveraging natural cooling,clean energy,and cost-effective resources.By 2030,this move is expected to reduce emissions from the data center sector by 16%–20%,generating direct economic benefits of approximately 53 billion USD.The success of this initiative can serve as a model for other countries to develop their internet infrastructure.展开更多
Open networks and heterogeneous services in the Internet of Vehicles(IoV)can lead to security and privacy challenges.One key requirement for such systems is the preservation of user privacy,ensuring a seamless experie...Open networks and heterogeneous services in the Internet of Vehicles(IoV)can lead to security and privacy challenges.One key requirement for such systems is the preservation of user privacy,ensuring a seamless experience in driving,navigation,and communication.These privacy needs are influenced by various factors,such as data collected at different intervals,trip durations,and user interactions.To address this,the paper proposes a Support Vector Machine(SVM)model designed to process large amounts of aggregated data and recommend privacy preserving measures.The model analyzes data based on user demands and interactions with service providers or neighboring infrastructure.It aims to minimize privacy risks while ensuring service continuity and sustainability.The SVMmodel helps validate the system’s reliability by creating a hyperplane that distinguishes between maximum and minimum privacy recommendations.The results demonstrate the effectiveness of the proposed SVM model in enhancing both privacy and service performance.展开更多
This study examines the effectiveness of artificial intelligence techniques in generating high-quality environmental data for species introductory site selection systems.Combining Strengths,Weaknesses,Opportunities,Th...This study examines the effectiveness of artificial intelligence techniques in generating high-quality environmental data for species introductory site selection systems.Combining Strengths,Weaknesses,Opportunities,Threats(SWOT)analysis data with Variation Autoencoder(VAE)and Generative AdversarialNetwork(GAN)the network framework model(SAE-GAN),is proposed for environmental data reconstruction.The model combines two popular generative models,GAN and VAE,to generate features conditional on categorical data embedding after SWOT Analysis.The model is capable of generating features that resemble real feature distributions and adding sample factors to more accurately track individual sample data.Reconstructed data is used to retain more semantic information to generate features.The model was applied to species in Southern California,USA,citing SWOT analysis data to train the model.Experiments show that the model is capable of integrating data from more comprehensive analyses than traditional methods and generating high-quality reconstructed data from them,effectively solving the problem of insufficient data collection in development environments.The model is further validated by the Technique for Order Preference by Similarity to an Ideal Solution(TOPSIS)classification assessment commonly used in the environmental data domain.This study provides a reliable and rich source of training data for species introduction site selection systems and makes a significant contribution to ecological and sustainable development.展开更多
Propelled by the rise of artificial intelligence,cloud services,and data center applications,next-generation,low-power,local-oscillator-less,digital signal processing(DSP)-free,and short-reach coherent optical communi...Propelled by the rise of artificial intelligence,cloud services,and data center applications,next-generation,low-power,local-oscillator-less,digital signal processing(DSP)-free,and short-reach coherent optical communication has evolved into an increasingly prominent area of research in recent years.Here,we demonstrate DSP-free coherent optical transmission by analog signal processing in frequency synchronous optical network(FSON)architecture,which supports polarization multiplexing and higher-order modulation formats.The FSON architecture that allows the numerous laser sources of optical transceivers within a data center can be quasi-synchronized by means of a tree-distributed homology architecture.In conjunction with our proposed pilot-tone assisted Costas loop for an analog coherent receiver,we achieve a record dual-polarization 224-Gb/s 16-QAM 5-km mismatch transmission with reset-free carrier phase recovery in the optical domain.Our proposed DSP-free analog coherent detection system based on the FSON makes it a promising solution for next-generation,low-power,and high-capacity coherent data center interconnects.展开更多
We study the two-dimensional(2D)Cauchy problem of nonhomogeneous Boussinesq system for magnetohydrodynamics convection without heat diffusion in the whole plane.Based on delicate weighted estimates,we derive the globa...We study the two-dimensional(2D)Cauchy problem of nonhomogeneous Boussinesq system for magnetohydrodynamics convection without heat diffusion in the whole plane.Based on delicate weighted estimates,we derive the global existence and uniqueness of strong solutions.In particular,the initial data can be arbitrarily large and the initial density may contain vacuum states and even have compact support.展开更多
The integration of image analysis through deep learning(DL)into rock classification represents a significant leap forward in geological research.While traditional methods remain invaluable for their expertise and hist...The integration of image analysis through deep learning(DL)into rock classification represents a significant leap forward in geological research.While traditional methods remain invaluable for their expertise and historical context,DL offers a powerful complement by enhancing the speed,objectivity,and precision of the classification process.This research explores the significance of image data augmentation techniques in optimizing the performance of convolutional neural networks(CNNs)for geological image analysis,particularly in the classification of igneous,metamorphic,and sedimentary rock types from rock thin section(RTS)images.This study primarily focuses on classic image augmentation techniques and evaluates their impact on model accuracy and precision.Results demonstrate that augmentation techniques like Equalize significantly enhance the model's classification capabilities,achieving an F1-Score of 0.9869 for igneous rocks,0.9884 for metamorphic rocks,and 0.9929 for sedimentary rocks,representing improvements compared to the baseline original results.Moreover,the weighted average F1-Score across all classes and techniques is 0.9886,indicating an enhancement.Conversely,methods like Distort lead to decreased accuracy and F1-Score,with an F1-Score of 0.949 for igneous rocks,0.954 for metamorphic rocks,and 0.9416 for sedimentary rocks,exacerbating the performance compared to the baseline.The study underscores the practicality of image data augmentation in geological image classification and advocates for the adoption of DL methods in this domain for automation and improved results.The findings of this study can benefit various fields,including remote sensing,mineral exploration,and environmental monitoring,by enhancing the accuracy of geological image analysis both for scientific research and industrial applications.展开更多
Handling missing data accurately is critical in clinical research, where data quality directly impacts decision-making and patient outcomes. While deep learning (DL) techniques for data imputation have gained attentio...Handling missing data accurately is critical in clinical research, where data quality directly impacts decision-making and patient outcomes. While deep learning (DL) techniques for data imputation have gained attention, challenges remain, especially when dealing with diverse data types. In this study, we introduce a novel data imputation method based on a modified convolutional neural network, specifically, a Deep Residual-Convolutional Neural Network (DRes-CNN) architecture designed to handle missing values across various datasets. Our approach demonstrates substantial improvements over existing imputation techniques by leveraging residual connections and optimized convolutional layers to capture complex data patterns. We evaluated the model on publicly available datasets, including Medical Information Mart for Intensive Care (MIMIC-III and MIMIC-IV), which contain critical care patient data, and the Beijing Multi-Site Air Quality dataset, which measures environmental air quality. The proposed DRes-CNN method achieved a root mean square error (RMSE) of 0.00006, highlighting its high accuracy and robustness. We also compared with Low Light-Convolutional Neural Network (LL-CNN) and U-Net methods, which had RMSE values of 0.00075 and 0.00073, respectively. This represented an improvement of approximately 92% over LL-CNN and 91% over U-Net. The results showed that this DRes-CNN-based imputation method outperforms current state-of-the-art models. These results established DRes-CNN as a reliable solution for addressing missing data.展开更多
In this paper,we propose a symmetric difference data enhancement physics-informed neural network(SDE-PINN)to study soliton solutions for discrete nonlinear lattice equations(NLEs).By considering known and unknown symm...In this paper,we propose a symmetric difference data enhancement physics-informed neural network(SDE-PINN)to study soliton solutions for discrete nonlinear lattice equations(NLEs).By considering known and unknown symmetric points,numerical simulations are conducted to one-soliton and two-soliton solutions of a discrete KdV equation,as well as a one-soliton solution of a discrete Toda lattice equation.Compared with the existing discrete deep learning approach,the numerical results reveal that within the specified spatiotemporal domain,the prediction accuracy by SDE-PINN is excellent regardless of the interior or extrapolation prediction,with a significant reduction in training time.The proposed data enhancement technique and symmetric structure development provides a new perspective for the deep learning approach to solve discrete NLEs.The newly proposed SDE-PINN can also be applied to solve continuous nonlinear equations and other discrete NLEs numerically.展开更多
基金supported by the School of Digital Science,Universiti Brunei Darussalam,Brunei.
文摘Artificial Intelligence(AI)in healthcare enables predicting diabetes using data-driven methods instead of the traditional ways of screening the disease,which include hemoglobin A1c(HbA1c),oral glucose tolerance test(OGTT),and fasting plasma glucose(FPG)screening techniques,which are invasive and limited in scale.Machine learning(ML)and deep neural network(DNN)models that use large datasets to learn the complex,nonlinear feature interactions,but the conventional ML algorithms are data sensitive and often show unstable predictive accuracy.Conversely,DNN models are more robust,though the ability to reach a high accuracy rate consistently on heterogeneous datasets is still an open challenge.For predicting diabetes,this work proposed a hybrid DNN approach by integrating a bidirectional long short-term memory(BiLSTM)network with a bidirectional gated recurrent unit(BiGRU).A robust DL model,developed by combining various datasets with weighted coefficients,dense operations in the connection of deep layers,and the output aggregation using batch normalization and dropout functions to avoid overfitting.The goal of this hybrid model is better generalization and consistency among various datasets,which facilitates the effective management and early intervention.The proposed DNN model exhibits an excellent predictive performance as compared to the state-of-the-art and baseline ML and DNN models for diabetes prediction tasks.The robust performance indicates the possible usefulness of DL-based models in the development of disease prediction in healthcare and other areas that demand high-quality analytics.
基金supported by the National Science Foundation of China(No.62171387)the Science and Technology Program of Sichuan Province(No.2024NSFSC0468)the China Postdoctoral Science Foundation(No.2019M663475).
文摘As an important resource in data link,time slots should be strategically allocated to enhance transmission efficiency and resist eavesdropping,especially considering the tremendous increase in the number of nodes and diverse communication needs.It is crucial to design control sequences with robust randomness and conflict-freeness to properly address differentiated access control in data link.In this paper,we propose a hierarchical access control scheme based on control sequences to achieve high utilization of time slots and differentiated access control.A theoretical bound of the hierarchical control sequence set is derived to characterize the constraints on the parameters of the sequence set.Moreover,two classes of optimal hierarchical control sequence sets satisfying the theoretical bound are constructed,both of which enable the scheme to achieve maximum utilization of time slots.Compared with the fixed time slot allocation scheme,our scheme reduces the symbol error rate by up to 9%,which indicates a significant improvement in anti-interference and eavesdropping capabilities.
基金financially supported by the National Science and Technology Major Project,China(No.2024ZD1002100)the National Natural Science Foundation of China(Nos.42330801,42474112,42504062)+1 种基金the China Postdoctoral Science Foundation(No.2024M761704)Shuimu Tsinghua Scholar Program of Tsinghua University,China(No.2024SM114)。
文摘A three-dimensional(3D)electromagnetic(EM)inversion algorithm based on the nonlinear conjugate gradient(NLCG)method and a two-color plane Gauss-Seidel(GS)multigrid(MG)forward solver is developed to improve inversion efficiency.The results indicate that the computational efficiency of each inversion can be improved by approximately a factor of three by using the proposed MG solver.First,the accuracy of the MG solver is validated through a test on a synthetic model.Next,the numerical performance of the inversion algorithm is evaluated using this model.Finally,the inversion algorithm is applied to a field EM data collected at the Beiya gold polymetallic ore district.A 3D resistivity model is obtained,and the formation process of the metal ore is analyzed.
基金supported by ZTE Industry-University-Institute Cooperation Funds under Grant No.HC-CN-20220607009。
文摘With the evolution of next-generation network technologies,the complexity of network management has significantly increased,and the means of network attacks are diversified,bringing new challenges to network traffic classification.This paper presents a general AIdriven network traffic classification workflow and elaborates on a traffic data and feature engineering framework.Most importantly,it analyzes the concept and causes of data distribution shifts in ne twork traffic,proposing detection methods and countermeasures.Experimental results on real traffic collected at different time intervals show that application evolution can induce data distribution shifts,which in turn lead to a noticeable degradation in traffic classification performance.Comparative drift detection experiments further confirm that such shifts are more evident over long-term intervals,while short-term traffic remains relatively stable.These findings demonstrate the necessity of incorporating drift-aware mechanisms into AI-driven network traffic classification systems.
文摘Indeed,a range of systems in the environment requires timely,spatially explicit,and credible information to support its environmental decision-making,but no one observing system can give the complete and reliable measures of the Earth system across scales.This review summarizes how the realization of the Compute the Planet is underway in the form of machine learning,remote sensing,and sensor data fusion to generate decision-ready environmental insights.We use the application-first approach,which considers remote sensing,in situ and Internet of Things(IoT)sensing,and physics-based models as complementary streams of evidence with similar strengths and failures.We look critically at how an integrated system can convert heterogeneous observations to action products across three high impact application areas:atmosphere and air quality,water–land–ecosystem dynamics,and hazards.Rapid-response situational awareness,ecosystem condition metrics,drought and flood indicators,exposure maps,and hazard/extreme indicators are key products.The integrated systems to environment interface in three high impact application areas:atmosphere and air quality,water-land-ecosystem dynamics,and hazard Examine Our operational requirements can often determine real-life value such as latency,time stability,smooth degradation in the presence of missing or degraded inputs,and calibrated uncertainty usable in thresholdbased decisions.These pitfalls are common across fields:mismatch in the scale between a point sensor and a gridded product,objectives on proxies in remotely sensed measurements,domain shift in the extremes and changing baselines,and evaluation aspects,which overestimate generalization because of spatiotemporal autocorrelation.Based on these lessons,we present cross-domain proposals for strong validation,uncertainty quantification,provenance,and versioning,as well as fair performance evaluation.We conclude that the next era of environmental intelligence will see a reduction in average accuracy improvement and an increase in terms of robustness,transparency,and operational responsibility,thus allowing the integrated environmental intelligence system to be deployed,which may be relied on to monitor human health,resource allocation,and survival in a more climate-adapted world.
基金supported by ZTE Industry-University-Institute Cooperation Funds under Grant No.IA20230628015the State Key Laboratory of Particle Detection and Electronics under Grant No.SKLPDE-KF-202314。
文摘Security and access control for data storage in 5G industrial Internet collaborative systems are facing significant challenges.The characteristics of 5 G networks,such as low latency and high speed,facilitate data transmission in the industrial Internet but also increase vulnerability to attacks like theft and tampering.Moreover,in 5G industrial Internet collaborative system environments,data flows across multiple entities and links,which necessitates a flexible access control model to meet specific data access requirements.Traditional role-based and attribute-based access control mechanisms are difficult to apply in such dynamic application scenarios.To address these challenges,we propose a novel data storage solution for 5G industrial Internet collaborative systems.Similar to existing approaches,it provides integrity and confidentiality protection for transmitted data.In terms of security,only authenticated data owners and users can obtain file decryption keys,preventing malicious attackers from data forgery.Regarding access control,decryption is permitted only to authorized data users,safeguarding against unauthorized file access.Furthermore,by introducing an attribute-based encryption mechanism,only data users with specific attributes can decrypt files.In terms of efficiency,our approach utilizes bilinear and modular exponentiation operations solely during the authentication process.For handling substantial data loads,lightweight cryptographic algorithms are employed.Consequently,our solution achieves higher efficiency compared with other known methods.Experimental results demonstrate the feasibility of our approach in real-world applications.
基金The National Natural Science Foundation of China(No.52361165658,U24A20169).
文摘A dual‑task parallel machine learning framework was developed by integrating a convolutional autoencoder(CAE)and a fully connected neural network(FCNN)via the gradient‑coupled mechanism,enabling simultaneous data compression‑reconstruction and structural damage identification.Under the condition where 40% of the sensor nodes are missing,the model successfully reconstructs the full sensor network with an R^(2) of 0.916 and normalized root mean square error(NRMSE)of 0.0288.Even under significant noise contamination with an SNR of 12 dB,the model maintains strong reconstruction performance,achieving a R^(2) of 0.910 and NRMSE of 0.0253.Forty‑six structural damage scenarios were simulated using the scaled bridge model.The accuracy of spatial localization and quantification of the damage severity using the framework exceeds 99.3%.The proposed framework reduces the training time by 54.4%and iteration counts by 45.5% compared to conventional two‑stage machine learning approaches,demonstrating the efficiency of gradient‑coupled optimization.
基金The work described in this paper was fully supported by a grant from Hong Kong Metropolitan University(RIF/2021/05).
文摘Parkinson’s disease(PD)is a debilitating neurological disorder affecting over 10 million people worldwide.PD classification models using voice signals as input are common in the literature.It is believed that using deep learning algorithms further enhances performance;nevertheless,it is challenging due to the nature of small-scale and imbalanced PD datasets.This paper proposed a convolutional neural network-based deep support vector machine(CNN-DSVM)to automate the feature extraction process using CNN and extend the conventional SVM to a DSVM for better classification performance in small-scale PD datasets.A customized kernel function reduces the impact of biased classification towards the majority class(healthy candidates in our consideration).An improved generative adversarial network(IGAN)was designed to generate additional training data to enhance the model’s performance.For performance evaluation,the proposed algorithm achieves a sensitivity of 97.6%and a specificity of 97.3%.The performance comparison is evaluated from five perspectives,including comparisons with different data generation algorithms,feature extraction techniques,kernel functions,and existing works.Results reveal the effectiveness of the IGAN algorithm,which improves the sensitivity and specificity by 4.05%–4.72%and 4.96%–5.86%,respectively;and the effectiveness of the CNN-DSVM algorithm,which improves the sensitivity by 1.24%–57.4%and specificity by 1.04%–163%and reduces biased detection towards the majority class.The ablation experiments confirm the effectiveness of individual components.Two future research directions have also been suggested.
基金supported by the National Natural Science Foundation of China under Grant No.61072061the 111 Project of China under Grant No.B08004the Fundamental Research Funds for the Central Universities under Grant No.2013RC0114
文摘Understanding the dynamic traffic and usage characteristics of data services in cellular networks is important for optimising network resources and improving user experience.Recent studies have illustrated traffic characteristics from specific perspectives,such as user behaviour,device type,and applications.In this paper,we present the results of our study from a different perspective,namely service providers,to reveal the traffic characteristics of cellular data networks.Our study is based on traffic data collected over a five-day period from a leading mobile operator's core network in China.We propose a Zipf-like model to characterise the distributions of the traffic volume,subscribers,and requests among service providers.Nine distinct diurnal traffic patterns of service providers are identified by formulating and solving a time series clustering problem.Our work differs from previous related works in that we perform measurements on a large quantity of data covering 2.2 billion traffic records,and we first explore the traffic patterns of thousands of service providers.Results of our study present mobile Internet participants with a better understanding of the traffic and usage characteristics of service providers,which play a critical role in the mobile Internet era.
基金funded by the Bavarian State Ministry of Science,Research and Art(Grant number:H.2-F1116.WE/52/2)。
文摘In order to address the widespread data shortage problem in battery research,this paper proposes a generative adversarial network model that combines it with deep convolutional networks,the Wasserstein distance,and the gradient penalty to achieve data augmentation.To lower the threshold for implementing the proposed method,transfer learning is further introduced.The W-DC-GAN-GP-TL framework is thereby formed.This framework is evaluated on 3 different publicly available datasets to judge the quality of generated data.Through visual comparisons and the examination of two visualization methods(probability density function(PDF)and principal component analysis(PCA)),it is demonstrated that the generated data is hard to distinguish from the real data.The application of generated data for training a battery state model using transfer learning is further evaluated.Specifically,Bi-GRU-based and Transformer-based methods are implemented on 2 separate datasets for estimating state of health(SOH)and state of charge(SOC),respectively.The results indicate that the proposed framework demonstrates satisfactory performance in different scenarios:for the data replacement scenario,where real data are removed and replaced with generated data,the state estimator accuracy decreases only slightly;for the data enhancement scenario,the estimator accuracy is further improved.The estimation accuracy of SOH and SOC is as low as 0.69%and 0.58%root mean square error(RMSE)after applying the proposed framework.This framework provides a reliable method for enriching battery measurement data.It is a generalized framework capable of generating a variety of time series data.
基金supported by the National Natural Science Foundation of China(42250101)the Macao Foundation。
文摘Earth’s internal core and crustal magnetic fields,as measured by geomagnetic satellites like MSS-1(Macao Science Satellite-1)and Swarm,are vital for understanding core dynamics and tectonic evolution.To model these internal magnetic fields accurately,data selection based on specific criteria is often employed to minimize the influence of rapidly changing current systems in the ionosphere and magnetosphere.However,the quantitative impact of various data selection criteria on internal geomagnetic field modeling is not well understood.This study aims to address this issue and provide a reference for constructing and applying geomagnetic field models.First,we collect the latest MSS-1 and Swarm satellite magnetic data and summarize widely used data selection criteria in geomagnetic field modeling.Second,we briefly describe the method to co-estimate the core,crustal,and large-scale magnetospheric fields using satellite magnetic data.Finally,we conduct a series of field modeling experiments with different data selection criteria to quantitatively estimate their influence.Our numerical experiments confirm that without selecting data from dark regions and geomagnetically quiet times,the resulting internal field differences at the Earth’s surface can range from tens to hundreds of nanotesla(nT).Additionally,we find that the uncertainties introduced into field models by different data selection criteria are significantly larger than the measurement accuracy of modern geomagnetic satellites.These uncertainties should be considered when utilizing constructed magnetic field models for scientific research and applications.
基金supported by the Joint Research Project for the Yangtze River Conservation(Phase II),China(2022-LHYJ-02-0401).
文摘As the world’s largest digital economy,China has a significant demand for data centers,which are energy-intensive.With an annual growth rate of 28%in installed capacity,these centers are primarily located in the developed eastern region,where land and energy resources are limited.This localization poses a major challenge to the industry’s net-zero goal.To address this,China has launched a bold initiative to relocate data centers to the western region,leveraging natural cooling,clean energy,and cost-effective resources.By 2030,this move is expected to reduce emissions from the data center sector by 16%–20%,generating direct economic benefits of approximately 53 billion USD.The success of this initiative can serve as a model for other countries to develop their internet infrastructure.
基金supported by the Deanship of Graduate Studies and Scientific Research at University of Bisha for funding this research through the promising program under grant number(UB-Promising-33-1445).
文摘Open networks and heterogeneous services in the Internet of Vehicles(IoV)can lead to security and privacy challenges.One key requirement for such systems is the preservation of user privacy,ensuring a seamless experience in driving,navigation,and communication.These privacy needs are influenced by various factors,such as data collected at different intervals,trip durations,and user interactions.To address this,the paper proposes a Support Vector Machine(SVM)model designed to process large amounts of aggregated data and recommend privacy preserving measures.The model analyzes data based on user demands and interactions with service providers or neighboring infrastructure.It aims to minimize privacy risks while ensuring service continuity and sustainability.The SVMmodel helps validate the system’s reliability by creating a hyperplane that distinguishes between maximum and minimum privacy recommendations.The results demonstrate the effectiveness of the proposed SVM model in enhancing both privacy and service performance.
基金supported by the Fundamental Research Funds for the Liaoning Universities(LJ212410146025).
文摘This study examines the effectiveness of artificial intelligence techniques in generating high-quality environmental data for species introductory site selection systems.Combining Strengths,Weaknesses,Opportunities,Threats(SWOT)analysis data with Variation Autoencoder(VAE)and Generative AdversarialNetwork(GAN)the network framework model(SAE-GAN),is proposed for environmental data reconstruction.The model combines two popular generative models,GAN and VAE,to generate features conditional on categorical data embedding after SWOT Analysis.The model is capable of generating features that resemble real feature distributions and adding sample factors to more accurately track individual sample data.Reconstructed data is used to retain more semantic information to generate features.The model was applied to species in Southern California,USA,citing SWOT analysis data to train the model.Experiments show that the model is capable of integrating data from more comprehensive analyses than traditional methods and generating high-quality reconstructed data from them,effectively solving the problem of insufficient data collection in development environments.The model is further validated by the Technique for Order Preference by Similarity to an Ideal Solution(TOPSIS)classification assessment commonly used in the environmental data domain.This study provides a reliable and rich source of training data for species introduction site selection systems and makes a significant contribution to ecological and sustainable development.
基金supported by the National Natural Science Foundation of China(Grant Nos.62405250 and 62471404)the China Postdoctoral Science Foundation(Grant No.2024M762955)+1 种基金the Key Project of Westlake Institute for Optoelectronics(Grant No.2023GD003)the Optical Com-munication and Sensing Laboratory,School of Engineering,Westlake University.
文摘Propelled by the rise of artificial intelligence,cloud services,and data center applications,next-generation,low-power,local-oscillator-less,digital signal processing(DSP)-free,and short-reach coherent optical communication has evolved into an increasingly prominent area of research in recent years.Here,we demonstrate DSP-free coherent optical transmission by analog signal processing in frequency synchronous optical network(FSON)architecture,which supports polarization multiplexing and higher-order modulation formats.The FSON architecture that allows the numerous laser sources of optical transceivers within a data center can be quasi-synchronized by means of a tree-distributed homology architecture.In conjunction with our proposed pilot-tone assisted Costas loop for an analog coherent receiver,we achieve a record dual-polarization 224-Gb/s 16-QAM 5-km mismatch transmission with reset-free carrier phase recovery in the optical domain.Our proposed DSP-free analog coherent detection system based on the FSON makes it a promising solution for next-generation,low-power,and high-capacity coherent data center interconnects.
文摘We study the two-dimensional(2D)Cauchy problem of nonhomogeneous Boussinesq system for magnetohydrodynamics convection without heat diffusion in the whole plane.Based on delicate weighted estimates,we derive the global existence and uniqueness of strong solutions.In particular,the initial data can be arbitrarily large and the initial density may contain vacuum states and even have compact support.
文摘The integration of image analysis through deep learning(DL)into rock classification represents a significant leap forward in geological research.While traditional methods remain invaluable for their expertise and historical context,DL offers a powerful complement by enhancing the speed,objectivity,and precision of the classification process.This research explores the significance of image data augmentation techniques in optimizing the performance of convolutional neural networks(CNNs)for geological image analysis,particularly in the classification of igneous,metamorphic,and sedimentary rock types from rock thin section(RTS)images.This study primarily focuses on classic image augmentation techniques and evaluates their impact on model accuracy and precision.Results demonstrate that augmentation techniques like Equalize significantly enhance the model's classification capabilities,achieving an F1-Score of 0.9869 for igneous rocks,0.9884 for metamorphic rocks,and 0.9929 for sedimentary rocks,representing improvements compared to the baseline original results.Moreover,the weighted average F1-Score across all classes and techniques is 0.9886,indicating an enhancement.Conversely,methods like Distort lead to decreased accuracy and F1-Score,with an F1-Score of 0.949 for igneous rocks,0.954 for metamorphic rocks,and 0.9416 for sedimentary rocks,exacerbating the performance compared to the baseline.The study underscores the practicality of image data augmentation in geological image classification and advocates for the adoption of DL methods in this domain for automation and improved results.The findings of this study can benefit various fields,including remote sensing,mineral exploration,and environmental monitoring,by enhancing the accuracy of geological image analysis both for scientific research and industrial applications.
基金supported by the Intelligent System Research Group(ISysRG)supported by Universitas Sriwijaya funded by the Competitive Research 2024.
文摘Handling missing data accurately is critical in clinical research, where data quality directly impacts decision-making and patient outcomes. While deep learning (DL) techniques for data imputation have gained attention, challenges remain, especially when dealing with diverse data types. In this study, we introduce a novel data imputation method based on a modified convolutional neural network, specifically, a Deep Residual-Convolutional Neural Network (DRes-CNN) architecture designed to handle missing values across various datasets. Our approach demonstrates substantial improvements over existing imputation techniques by leveraging residual connections and optimized convolutional layers to capture complex data patterns. We evaluated the model on publicly available datasets, including Medical Information Mart for Intensive Care (MIMIC-III and MIMIC-IV), which contain critical care patient data, and the Beijing Multi-Site Air Quality dataset, which measures environmental air quality. The proposed DRes-CNN method achieved a root mean square error (RMSE) of 0.00006, highlighting its high accuracy and robustness. We also compared with Low Light-Convolutional Neural Network (LL-CNN) and U-Net methods, which had RMSE values of 0.00075 and 0.00073, respectively. This represented an improvement of approximately 92% over LL-CNN and 91% over U-Net. The results showed that this DRes-CNN-based imputation method outperforms current state-of-the-art models. These results established DRes-CNN as a reliable solution for addressing missing data.
基金supported by the National Natural Science Foundation of China(Grant No.12071042)the Beijing Natural Science Foundation(Grant No.1202004)Promoting the Development of University Classification-Student Innovation and Entrepreneurship Training Programme(Grant No.5112410857)。
文摘In this paper,we propose a symmetric difference data enhancement physics-informed neural network(SDE-PINN)to study soliton solutions for discrete nonlinear lattice equations(NLEs).By considering known and unknown symmetric points,numerical simulations are conducted to one-soliton and two-soliton solutions of a discrete KdV equation,as well as a one-soliton solution of a discrete Toda lattice equation.Compared with the existing discrete deep learning approach,the numerical results reveal that within the specified spatiotemporal domain,the prediction accuracy by SDE-PINN is excellent regardless of the interior or extrapolation prediction,with a significant reduction in training time.The proposed data enhancement technique and symmetric structure development provides a new perspective for the deep learning approach to solve discrete NLEs.The newly proposed SDE-PINN can also be applied to solve continuous nonlinear equations and other discrete NLEs numerically.
