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
Accurate purchase prediction in e-commerce critically depends on the quality of behavioral features.This paper proposes a layered and interpretable feature engineering framework that organizes user signals into three ...Accurate purchase prediction in e-commerce critically depends on the quality of behavioral features.This paper proposes a layered and interpretable feature engineering framework that organizes user signals into three layers:Basic,Conversion&Stability(efficiency and volatility across actions),and Advanced Interactions&Activity(crossbehavior synergies and intensity).Using real Taobao(Alibaba’s primary e-commerce platform)logs(57,976 records for 10,203 users;25 November–03 December 2017),we conducted a hierarchical,layer-wise evaluation that holds data splits and hyperparameters fixed while varying only the feature set to quantify each layer’s marginal contribution.Across logistic regression(LR),decision tree,random forest,XGBoost,and CatBoost models with stratified 5-fold cross-validation,the performance improvedmonotonically fromBasic to Conversion&Stability to Advanced features.With LR,F1 increased from 0.613(Basic)to 0.962(Advanced);boosted models achieved high discrimination(0.995 AUC Score)and an F1 score up to 0.983.Calibration and precision–recall analyses indicated strong ranking quality and acknowledged potential dataset and period biases given the short(9-day)window.By making feature contributions measurable and reproducible,the framework complements model-centric advances and offers a transparent blueprint for production-grade behavioralmodeling.The code and processed artifacts are publicly available,and future work will extend the validation to longer,seasonal datasets and hybrid approaches that combine automated feature learning with domain-driven design.展开更多
Small datasets are often challenging due to their limited sample size.This research introduces a novel solution to these problems:average linkage virtual sample generation(ALVSG).ALVSG leverages the underlying data st...Small datasets are often challenging due to their limited sample size.This research introduces a novel solution to these problems:average linkage virtual sample generation(ALVSG).ALVSG leverages the underlying data structure to create virtual samples,which can be used to augment the original dataset.The ALVSG process consists of two steps.First,an average-linkage clustering technique is applied to the dataset to create a dendrogram.The dendrogram represents the hierarchical structure of the dataset,with each merging operation regarded as a linkage.Next,the linkages are combined into an average-based dataset,which serves as a new representation of the dataset.The second step in the ALVSG process involves generating virtual samples using the average-based dataset.The research project generates a set of 100 virtual samples by uniformly distributing them within the provided boundary.These virtual samples are then added to the original dataset,creating a more extensive dataset with improved generalization performance.The efficacy of the ALVSG approach is validated through resampling experiments and t-tests conducted on two small real-world datasets.The experiments are conducted on three forecasting models:the support vector machine for regression(SVR),the deep learning model(DL),and XGBoost.The results show that the ALVSG approach outperforms the baseline methods in terms of mean square error(MSE),root mean square error(RMSE),and mean absolute error(MAE).展开更多
With the deep integration of smart manufacturing and IoT technologies,higher demands are placed on the intelligence and real-time performance of industrial equipment fault detection.For industrial fans,base bolt loose...With the deep integration of smart manufacturing and IoT technologies,higher demands are placed on the intelligence and real-time performance of industrial equipment fault detection.For industrial fans,base bolt loosening faults are difficult to identify through conventional spectrum analysis,and the extreme scarcity of fault data leads to limited training datasets,making traditional deep learning methods inaccurate in fault identification and incapable of detecting loosening severity.This paper employs Bayesian Learning by training on a small fault dataset collected from the actual operation of axial-flow fans in a factory to obtain posterior distribution.This method proposes specific data processing approaches and a configuration of Bayesian Convolutional Neural Network(BCNN).It can effectively improve the model’s generalization ability.Experimental results demonstrate high detection accuracy and alignment with real-world applications,offering practical significance and reference value for industrial fan bolt loosening detection under data-limited conditions.展开更多
Objective To investigate methods for constructing a high-quality instructional dataset for traditional Chinese medicine(TCM)mental disorders and to validate its efficacy.Methods We proposed the Fine-Med-Mental-T&P...Objective To investigate methods for constructing a high-quality instructional dataset for traditional Chinese medicine(TCM)mental disorders and to validate its efficacy.Methods We proposed the Fine-Med-Mental-T&P methodology for constructing high-quality instruction datasets in TCM mental disorders.This approach integrates theoretical knowledge and practical case studies through a dual-track strategy.(i)Theoretical track:textbooks and guidelines on TCM mental disorders were manually segmented.Initial responses were generated using DeepSeek-V3,followed by refinement by the Qwen3-32B model to align the expression with human preferences.A screening algorithm was then applied to select 16000 high-quality instruction pairs.(ii)Practical track:starting from over 600 real clinical case seeds,diagnostic and therapeutic instruction pairs were generated using DeepSeek-V3 and subsequently screened through manual evaluation,resulting in 4000 high-quality practiceoriented instruction pairs.The integration of both tracks yielded the Med-Mental-Instruct-T&P dataset,comprising a total of 20000 instruction pairs.To validate the dataset’s effectiveness,three experimental evaluations(both manual and automated)were conducted:(i)comparative studies to compare the performance of models fine-tuned on different datasets;(ii)benchmarking to compare against mainstream TCM-specific large language models(LLMs);(iii)data ablation study to investigate the relationship between data volume and model performance.Results Experimental results demonstrate the superior performance of T&P-model finetuned on the Med-Mental-Instruct-T&P dataset.In the comparative study,the T&P-model significantly outperformed the baseline models trained solely on self-generated or purely human-curated baseline data.This superiority was evident in both automated metrics(ROUGEL>0.55)and expert manual evaluations(scoring above 7/10 across accuracy).In benchmark comparisons,the T&P-model also excelled against existing mainstream TCM LLMs(e.g.,HuatuoGPT and ZuoyiGPT).It showed particularly strong capabilities in handling diverse clinical presentations,including challenging disorders such as insomnia and coma,showcasing its robustness and versatility.Data ablation studies showed that T&P-model performance had an overall upward trend with minor fluctuations when training data increased from 10%to 50%;beyond 50%,performance improvement slowed significantly,with metrics plateauing and approaching a saturation point.展开更多
Dear Editor,This letter presents techniques to simplify dataset generation for instance segmentation of raw meat products,a critical step toward automating food production lines.Accurate segmentation is essential for ...Dear Editor,This letter presents techniques to simplify dataset generation for instance segmentation of raw meat products,a critical step toward automating food production lines.Accurate segmentation is essential for addressing challenges such as occlusions,indistinct edges,and stacked configurations,which demand large,diverse datasets.To meet these demands,we propose two complementary approaches:a semi-automatic annotation interface using tools like the segment anything model(SAM)and GrabCut and a synthetic data generation pipeline leveraging 3D-scanned models.These methods reduce reliance on real meat,mitigate food waste,and improve scalability.Experimental results demonstrate that incorporating synthetic data enhances segmentation model performance and,when combined with real data,further boosts accuracy,paving the way for more efficient automation in the food industry.展开更多
Knowing the influence of the size of datasets for regression models can help in improving the accuracy of a solar power forecast and make the most out of renewable energy systems.This research explores the influence o...Knowing the influence of the size of datasets for regression models can help in improving the accuracy of a solar power forecast and make the most out of renewable energy systems.This research explores the influence of dataset size on the accuracy and reliability of regression models for solar power prediction,contributing to better forecasting methods.The study analyzes data from two solar panels,aSiMicro03036 and aSiTandem72-46,over 7,14,17,21,28,and 38 days,with each dataset comprising five independent and one dependent parameter,and split 80–20 for training and testing.Results indicate that Random Forest consistently outperforms other models,achieving the highest correlation coefficient of 0.9822 and the lowest Mean Absolute Error(MAE)of 2.0544 on the aSiTandem72-46 panel with 21 days of data.For the aSiMicro03036 panel,the best MAE of 4.2978 was reached using the k-Nearest Neighbor(k-NN)algorithm,which was set up as instance-based k-Nearest neighbors(IBk)in Weka after being trained on 17 days of data.Regression performance for most models(excluding IBk)stabilizes at 14 days or more.Compared to the 7-day dataset,increasing to 21 days reduced the MAE by around 20%and improved correlation coefficients by around 2.1%,highlighting the value of moderate dataset expansion.These findings suggest that datasets spanning 17 to 21 days,with 80%used for training,can significantly enhance the predictive accuracy of solar power generation models.展开更多
Scientific knowledge on the chemical compositions of fine particulate matter(PM_(2.5)) is essential for properly assessing its health and climate effects,and for decisionmakers to develop efficient mitigation strategi...Scientific knowledge on the chemical compositions of fine particulate matter(PM_(2.5)) is essential for properly assessing its health and climate effects,and for decisionmakers to develop efficient mitigation strategies.A high-resolution PM_(2.5) chemical composition dataset(CAQRA-aerosol)is developed in this study,which provides hourly maps of organic carbon,black carbon,ammonium,nitrate,and sulfate in China from 2013 to 2020 with a horizontal resolution of 15 km.This paper describes the method,access,and validation results of this dataset.It shows that CAQRA-aerosol has good consistency with observations and achieves higher or comparable accuracy with previous PM_(2.5) composition datasets.Based on CAQRA-aerosol,spatiotemporal changes of different PM_(2.5) compositions were investigated from a national viewpoint,which emphasizes different changes of nitrate from other compositions.The estimated annual rate of population-weighted concentrations of nitrate is 0.23μg m^(−3)yr^(−1) from 2015 to 2020,compared with−0.19 to−1.1μg m^(−3)yr^(−1) for other compositions.The whole dataset is freely available from the China Air Pollution Data Center(https://doi.org/10.12423/capdb_PKU.2023.DA).展开更多
Accurate fine-grained geospatial scene classification using remote sensing imagery is essential for a wide range of applications.However,existing approaches often rely on manually zooming remote sensing images at diff...Accurate fine-grained geospatial scene classification using remote sensing imagery is essential for a wide range of applications.However,existing approaches often rely on manually zooming remote sensing images at different scales to create typical scene samples.This approach fails to adequately support the fixed-resolution image interpretation requirements in real-world scenarios.To address this limitation,we introduce the million-scale fine-grained geospatial scene classification dataset(MEET),which contains over 1.03 million zoom-free remote sensing scene samples,manually annotated into 80 fine-grained categories.In MEET,each scene sample follows a scene-in-scene layout,where the central scene serves as the reference,and auxiliary scenes provide crucial spatial context for fine-grained classification.Moreover,to tackle the emerging challenge of scene-in-scene classification,we present the context-aware transformer(CAT),a model specifically designed for this task,which adaptively fuses spatial context to accurately classify the scene samples.CAT adaptively fuses spatial context to accurately classify the scene samples by learning attentional features that capture the relationships between the center and auxiliary scenes.Based on MEET,we establish a comprehensive benchmark for fine-grained geospatial scene classification,evaluating CAT against 11 competitive baselines.The results demonstrate that CAT significantly outperforms these baselines,achieving a 1.88%higher balanced accuracy(BA)with the Swin-Large backbone,and a notable 7.87%improvement with the Swin-Huge backbone.Further experiments validate the effectiveness of each module in CAT and show the practical applicability of CAT in the urban functional zone mapping.The source code and dataset will be publicly available at https://jerrywyn.github.io/project/MEET.html.展开更多
This data set collects,compares and contrasts the capacities and structures of a series of hard carbon materials,and then searches for correlations between structure and electrochemical performance.The capacity data o...This data set collects,compares and contrasts the capacities and structures of a series of hard carbon materials,and then searches for correlations between structure and electrochemical performance.The capacity data of the hard carbons were obtained by charge/discharge tests and the materials were characterized by XRD,gas adsorption,true density tests and SAXS.In particular,the fitting of SAXS gave a series of structural parameters which showed good characterization.The related test details are given with the structural data of the hard carbons and the electrochemical performance of the sodium-ion batteries.展开更多
The increasing adoption of Industrial Internet of Things(IIoT)systems in smart manufacturing is leading to raise cyberattack numbers and pressing the requirement for intrusion detection systems(IDS)to be effective.How...The increasing adoption of Industrial Internet of Things(IIoT)systems in smart manufacturing is leading to raise cyberattack numbers and pressing the requirement for intrusion detection systems(IDS)to be effective.However,existing datasets for IDS training often lack relevance to modern IIoT environments,limiting their applicability for research and development.To address the latter gap,this paper introduces the HiTar-2024 dataset specifically designed for IIoT systems.As a consequence,that can be used by an IDS to detect imminent threats.Likewise,HiTar-2024 was generated using the AREZZO simulator,which replicates realistic smart manufacturing scenarios.The generated dataset includes five distinct classes:Normal,Probing,Remote to Local(R2L),User to Root(U2R),and Denial of Service(DoS).Furthermore,comprehensive experiments with popular Machine Learning(ML)models using various classifiers,including BayesNet,Logistic,IBK,Multiclass,PART,and J48 demonstrate high accuracy,precision,recall,and F1-scores,exceeding 0.99 across all ML metrics.The latter result is reached thanks to the rigorous applied process to achieve this quite good result,including data pre-processing,features extraction,fixing the class imbalance problem,and using a test option for model robustness.This comprehensive approach emphasizes meticulous dataset construction through a complete dataset generation process,a careful labelling algorithm,and a sophisticated evaluation method,providing valuable insights to reinforce IIoT system security.Finally,the HiTar-2024 dataset is compared with other similar datasets in the literature,considering several factors such as data format,feature extraction tools,number of features,attack categories,number of instances,and ML metrics.展开更多
This paper introduces a robust Distributed Denial-of-Service attack detection framework tailored for Software-Defined Networking based Internet of Things environments,built upon a novel,syntheticmulti-vector dataset g...This paper introduces a robust Distributed Denial-of-Service attack detection framework tailored for Software-Defined Networking based Internet of Things environments,built upon a novel,syntheticmulti-vector dataset generated in a Mininet-Ryu testbed using real-time flow-based labeling.The proposed model is based on the XGBoost algorithm,optimized with Principal Component Analysis for dimensionality reduction,utilizing lightweight flowlevel features extracted from Open Flow statistics to classify attacks across critical IoT protocols including TCP,UDP,HTTP,MQTT,and CoAP.The model employs lightweight flow-level features extracted from Open Flow statistics to ensure low computational overhead and fast processing.Performance was rigorously evaluated using key metrics,including Accuracy,Precision,Recall,F1-Score,False Alarm Rate,AUC-ROC,and Detection Time.Experimental results demonstrate the model’s high performance,achieving an accuracy of 98.93%and a low FAR of 0.86%,with a rapid median detection time of 1.02 s.This efficiency validates its superiority in meeting critical Key Performance Indicators,such as Latency and high Throughput,necessary for time-sensitive SDN-IoT systems.Furthermore,the model’s robustness and statistically significant outperformance against baseline models such as Random Forest,k-Nearest Neighbors,and Gradient Boosting Machine,validating through statistical tests using Wilcoxon signed-rank test and confirmed via successful deployment in a real SDN testbed for live traffic detection and mitigation.展开更多
When dealing with imbalanced datasets,the traditional support vectormachine(SVM)tends to produce a classification hyperplane that is biased towards the majority class,which exhibits poor robustness.This paper proposes...When dealing with imbalanced datasets,the traditional support vectormachine(SVM)tends to produce a classification hyperplane that is biased towards the majority class,which exhibits poor robustness.This paper proposes a high-performance classification algorithm specifically designed for imbalanced datasets.The proposed method first uses a biased second-order cone programming support vectormachine(B-SOCP-SVM)to identify the support vectors(SVs)and non-support vectors(NSVs)in the imbalanced data.Then,it applies the synthetic minority over-sampling technique(SV-SMOTE)to oversample the support vectors of the minority class and uses the random under-sampling technique(NSV-RUS)multiple times to undersample the non-support vectors of the majority class.Combining the above-obtained minority class data set withmultiple majority class datasets can obtainmultiple new balanced data sets.Finally,SOCP-SVM is used to classify each data set,and the final result is obtained through the integrated algorithm.Experimental results demonstrate that the proposed method performs excellently on imbalanced datasets.展开更多
Detecting faces under occlusion remains a significant challenge in computer vision due to variations caused by masks,sunglasses,and other obstructions.Addressing this issue is crucial for applications such as surveill...Detecting faces under occlusion remains a significant challenge in computer vision due to variations caused by masks,sunglasses,and other obstructions.Addressing this issue is crucial for applications such as surveillance,biometric authentication,and human-computer interaction.This paper provides a comprehensive review of face detection techniques developed to handle occluded faces.Studies are categorized into four main approaches:feature-based,machine learning-based,deep learning-based,and hybrid methods.We analyzed state-of-the-art studies within each category,examining their methodologies,strengths,and limitations based on widely used benchmark datasets,highlighting their adaptability to partial and severe occlusions.The review also identifies key challenges,including dataset diversity,model generalization,and computational efficiency.Our findings reveal that deep learning methods dominate recent studies,benefiting from their ability to extract hierarchical features and handle complex occlusion patterns.More recently,researchers have increasingly explored Transformer-based architectures,such as Vision Transformer(ViT)and Swin Transformer,to further improve detection robustness under challenging occlusion scenarios.In addition,hybrid approaches,which aim to combine traditional andmodern techniques,are emerging as a promising direction for improving robustness.This review provides valuable insights for researchers aiming to develop more robust face detection systems and for practitioners seeking to deploy reliable solutions in real-world,occlusionprone environments.Further improvements and the proposal of broader datasets are required to developmore scalable,robust,and efficient models that can handle complex occlusions in real-world scenarios.展开更多
Climate change significantly affects environment,ecosystems,communities,and economies.These impacts often result in quick and gradual changes in water resources,environmental conditions,and weather patterns.A geograph...Climate change significantly affects environment,ecosystems,communities,and economies.These impacts often result in quick and gradual changes in water resources,environmental conditions,and weather patterns.A geographical study was conducted in Arizona State,USA,to examine monthly precipi-tation concentration rates over time.This analysis used a high-resolution 0.50×0.50 grid for monthly precip-itation data from 1961 to 2022,Provided by the Climatic Research Unit.The study aimed to analyze climatic changes affected the first and last five years of each decade,as well as the entire decade,during the specified period.GIS was used to meet the objectives of this study.Arizona experienced 51–568 mm,67–560 mm,63–622 mm,and 52–590 mm of rainfall in the sixth,seventh,eighth,and ninth decades of the second millennium,respectively.Both the first and second five year periods of each decade showed accept-able rainfall amounts despite fluctuations.However,rainfall decreased in the first and second decades of the third millennium.and in the first two years of the third decade.Rainfall amounts dropped to 42–472 mm,55–469 mm,and 74–498 mm,respectively,indicating a downward trend in precipitation.The central part of the state received the highest rainfall,while the eastern and western regions(spanning north to south)had significantly less.Over the decades of the third millennium,the average annual rainfall every five years was relatively low,showing a declining trend due to severe climate changes,generally ranging between 35 mm and 498 mm.The central regions consistently received more rainfall than the eastern and western outskirts.Arizona is currently experiencing a decrease in rainfall due to climate change,a situation that could deterio-rate further.This highlights the need to optimize the use of existing rainfall and explore alternative water sources.展开更多
Standardized datasets are foundational to healthcare informatization by enhancing data quality and unleashing the value of data elements.Using bibliometrics and content analysis,this study examines China's healthc...Standardized datasets are foundational to healthcare informatization by enhancing data quality and unleashing the value of data elements.Using bibliometrics and content analysis,this study examines China's healthcare dataset standards from 2011 to 2025.It analyzes their evolution across types,applications,institutions,and themes,highlighting key achievements including substantial growth in quantity,optimized typology,expansion into innovative application scenarios such as health decision support,and broadened institutional involvement.The study also identifies critical challenges,including imbalanced development,insufficient quality control,and a lack of essential metadata—such as authoritative data element mappings and privacy annotations—which hampers the delivery of intelligent services.To address these challenges,the study proposes a multi-faceted strategy focused on optimizing the standard system's architecture,enhancing quality and implementation,and advancing both data governance—through authoritative tracing and privacy protection—and intelligent service provision.These strategies aim to promote the application of dataset standards,thereby fostering and securing the development of new productive forces in healthcare.展开更多
Lithology identification is a critical aspect of geoenergy exploration,including geothermal energy development,gas hydrate extraction,and gas storage.In recent years,artificial intelligence techniques based on drill c...Lithology identification is a critical aspect of geoenergy exploration,including geothermal energy development,gas hydrate extraction,and gas storage.In recent years,artificial intelligence techniques based on drill core images have made significant strides in lithology identification,achieving high accuracy.However,the current demand for advanced lithology identification models remains unmet due to the lack of high-quality drill core image datasets.This study successfully constructs and publicly releases the first open-source Drill Core Image Dataset(DCID),addressing the need for large-scale,high-quality datasets in lithology characterization tasks within geological engineering and establishing a standard dataset for model evaluation.DCID consists of 35 lithology categories and a total of 98,000 high-resolution images(512×512 pixels),making it the most comprehensive drill core image dataset in terms of lithology categories,image quantity,and resolution.This study also provides lithology identification accuracy benchmarks for popular convolutional neural networks(CNNs)such as VGG,ResNet,DenseNet,MobileNet,as well as for the Vision Transformer(ViT)and MLP-Mixer,based on DCID.Additionally,the sensitivity of model performance to various parameters and image resolution is evaluated.In response to real-world challenges,we propose a real-world data augmentation(RWDA)method,leveraging slightly defective images from DCID to enhance model robustness.The study also explores the impact of real-world lighting conditions on the performance of lithology identification models.Finally,we demonstrate how to rapidly evaluate model performance across multiple dimensions using low-resolution datasets,advancing the application and development of new lithology identification models for geoenergy exploration.展开更多
Background:The life-course management of children with tetralogy of Fallot(TOF)has focused on demonstrating brain structural alterations,developmental trajectories,and cognition-related changes that unfold over time.M...Background:The life-course management of children with tetralogy of Fallot(TOF)has focused on demonstrating brain structural alterations,developmental trajectories,and cognition-related changes that unfold over time.Methods:We introduce an magnetic resonance imaging(MRI)dataset comprising TOF children who underwent brain MRI scanning and cross-sectional neurocognitive follow-up.The dataset includes brain three-dimensional T1-weighted imaging(3D-T1WI),three-dimensional T2-weighted imaging(3D-T2WI),and neurodevelopmental evaluations using the Wechsler Preschool and Primary Scale of Intelligence–Fourth Edition(WPPSI-IV).Results:Thirty-one children with TOF(age range:4–33 months;18 males)were recruited and completed corrective surgery at the Children’s Hospital of Nanjing Medical University,Nanjing,China.Aiming to promote the neurodevelopmental outcomes in children with TOF,we have meticulously curated a comprehensive dataset designed to dissect the complex interplay among risk factors,neuroimaging findings,and adverse neurodevelopmental outcomes.Conclusion:This article aims to introduce our open-source dataset on neurodevelopment in children with TOF,which covers the data types,data acquisition and processing methods,the procedure for accessing the data,and related publications.展开更多
The aim of this article is to explore potential directions for the development of artificial intelligence(AI).It points out that,while current AI can handle the statistical properties of complex systems,it has difficu...The aim of this article is to explore potential directions for the development of artificial intelligence(AI).It points out that,while current AI can handle the statistical properties of complex systems,it has difficulty effectively processing and fully representing their spatiotemporal complexity patterns.The article also discusses a potential path of AI development in the engineering domain.Based on the existing understanding of the principles of multilevel com-plexity,this article suggests that consistency among the logical structures of datasets,AI models,model-building software,and hardware will be an important AI development direction and is worthy of careful consideration.展开更多
The unprecedented scale of large models,such as large language models(LLMs)and text-to-image diffusion models,has raised critical concerns about the unauthorized use of copyrighted data during model training.These con...The unprecedented scale of large models,such as large language models(LLMs)and text-to-image diffusion models,has raised critical concerns about the unauthorized use of copyrighted data during model training.These concerns have spurred a growing demand for dataset copyright auditing techniques,which aim to detect and verify potential infringements in the training data of commercial AI systems.This paper presents a survey of existing auditing solutions,categorizing them across key dimensions:data modality,model training stage,data overlap scenarios,and model access levels.We highlight major trends,including the prevalence of black-box auditing methods and the emphasis on fine-tuning rather than pre-training.Through an in-depth analysis of 12 representative works,we extract four key observations that reveal the limitations of current methods.Furthermore,we identify three open challenges and propose future directions for robust,multimodal,and scalable auditing solutions.Our findings underscore the urgent need to establish standardized benchmarks and develop auditing frameworks that are resilient to low watermark densities and applicable in diverse deployment settings.展开更多
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 research fund of Hanyang University(HY-202500000001616).
文摘Accurate purchase prediction in e-commerce critically depends on the quality of behavioral features.This paper proposes a layered and interpretable feature engineering framework that organizes user signals into three layers:Basic,Conversion&Stability(efficiency and volatility across actions),and Advanced Interactions&Activity(crossbehavior synergies and intensity).Using real Taobao(Alibaba’s primary e-commerce platform)logs(57,976 records for 10,203 users;25 November–03 December 2017),we conducted a hierarchical,layer-wise evaluation that holds data splits and hyperparameters fixed while varying only the feature set to quantify each layer’s marginal contribution.Across logistic regression(LR),decision tree,random forest,XGBoost,and CatBoost models with stratified 5-fold cross-validation,the performance improvedmonotonically fromBasic to Conversion&Stability to Advanced features.With LR,F1 increased from 0.613(Basic)to 0.962(Advanced);boosted models achieved high discrimination(0.995 AUC Score)and an F1 score up to 0.983.Calibration and precision–recall analyses indicated strong ranking quality and acknowledged potential dataset and period biases given the short(9-day)window.By making feature contributions measurable and reproducible,the framework complements model-centric advances and offers a transparent blueprint for production-grade behavioralmodeling.The code and processed artifacts are publicly available,and future work will extend the validation to longer,seasonal datasets and hybrid approaches that combine automated feature learning with domain-driven design.
基金funding support from the National Science and Technology Council(NSTC),under Grant No.114-2410-H-011-026-MY3.
文摘Small datasets are often challenging due to their limited sample size.This research introduces a novel solution to these problems:average linkage virtual sample generation(ALVSG).ALVSG leverages the underlying data structure to create virtual samples,which can be used to augment the original dataset.The ALVSG process consists of two steps.First,an average-linkage clustering technique is applied to the dataset to create a dendrogram.The dendrogram represents the hierarchical structure of the dataset,with each merging operation regarded as a linkage.Next,the linkages are combined into an average-based dataset,which serves as a new representation of the dataset.The second step in the ALVSG process involves generating virtual samples using the average-based dataset.The research project generates a set of 100 virtual samples by uniformly distributing them within the provided boundary.These virtual samples are then added to the original dataset,creating a more extensive dataset with improved generalization performance.The efficacy of the ALVSG approach is validated through resampling experiments and t-tests conducted on two small real-world datasets.The experiments are conducted on three forecasting models:the support vector machine for regression(SVR),the deep learning model(DL),and XGBoost.The results show that the ALVSG approach outperforms the baseline methods in terms of mean square error(MSE),root mean square error(RMSE),and mean absolute error(MAE).
基金funded by the Zhejiang Provincial Key Science and Technology“LingYan”Project Foundation,grant number 2023C01145Zhejiang Gongshang University Higher Education Research Projects,grant number Xgy22028.
文摘With the deep integration of smart manufacturing and IoT technologies,higher demands are placed on the intelligence and real-time performance of industrial equipment fault detection.For industrial fans,base bolt loosening faults are difficult to identify through conventional spectrum analysis,and the extreme scarcity of fault data leads to limited training datasets,making traditional deep learning methods inaccurate in fault identification and incapable of detecting loosening severity.This paper employs Bayesian Learning by training on a small fault dataset collected from the actual operation of axial-flow fans in a factory to obtain posterior distribution.This method proposes specific data processing approaches and a configuration of Bayesian Convolutional Neural Network(BCNN).It can effectively improve the model’s generalization ability.Experimental results demonstrate high detection accuracy and alignment with real-world applications,offering practical significance and reference value for industrial fan bolt loosening detection under data-limited conditions.
基金Key Scientific Research Project of the Hunan Provincial Department of Education(23A312).
文摘Objective To investigate methods for constructing a high-quality instructional dataset for traditional Chinese medicine(TCM)mental disorders and to validate its efficacy.Methods We proposed the Fine-Med-Mental-T&P methodology for constructing high-quality instruction datasets in TCM mental disorders.This approach integrates theoretical knowledge and practical case studies through a dual-track strategy.(i)Theoretical track:textbooks and guidelines on TCM mental disorders were manually segmented.Initial responses were generated using DeepSeek-V3,followed by refinement by the Qwen3-32B model to align the expression with human preferences.A screening algorithm was then applied to select 16000 high-quality instruction pairs.(ii)Practical track:starting from over 600 real clinical case seeds,diagnostic and therapeutic instruction pairs were generated using DeepSeek-V3 and subsequently screened through manual evaluation,resulting in 4000 high-quality practiceoriented instruction pairs.The integration of both tracks yielded the Med-Mental-Instruct-T&P dataset,comprising a total of 20000 instruction pairs.To validate the dataset’s effectiveness,three experimental evaluations(both manual and automated)were conducted:(i)comparative studies to compare the performance of models fine-tuned on different datasets;(ii)benchmarking to compare against mainstream TCM-specific large language models(LLMs);(iii)data ablation study to investigate the relationship between data volume and model performance.Results Experimental results demonstrate the superior performance of T&P-model finetuned on the Med-Mental-Instruct-T&P dataset.In the comparative study,the T&P-model significantly outperformed the baseline models trained solely on self-generated or purely human-curated baseline data.This superiority was evident in both automated metrics(ROUGEL>0.55)and expert manual evaluations(scoring above 7/10 across accuracy).In benchmark comparisons,the T&P-model also excelled against existing mainstream TCM LLMs(e.g.,HuatuoGPT and ZuoyiGPT).It showed particularly strong capabilities in handling diverse clinical presentations,including challenging disorders such as insomnia and coma,showcasing its robustness and versatility.Data ablation studies showed that T&P-model performance had an overall upward trend with minor fluctuations when training data increased from 10%to 50%;beyond 50%,performance improvement slowed significantly,with metrics plateauing and approaching a saturation point.
基金supported by European Union’s Horizon Europe research and innovation programme,project AGILEHAND(Smart Grading,Handling and Packaging Solutions for Soft and Deformable Products in Agile and Reconfigurable Lines)(101092043).
文摘Dear Editor,This letter presents techniques to simplify dataset generation for instance segmentation of raw meat products,a critical step toward automating food production lines.Accurate segmentation is essential for addressing challenges such as occlusions,indistinct edges,and stacked configurations,which demand large,diverse datasets.To meet these demands,we propose two complementary approaches:a semi-automatic annotation interface using tools like the segment anything model(SAM)and GrabCut and a synthetic data generation pipeline leveraging 3D-scanned models.These methods reduce reliance on real meat,mitigate food waste,and improve scalability.Experimental results demonstrate that incorporating synthetic data enhances segmentation model performance and,when combined with real data,further boosts accuracy,paving the way for more efficient automation in the food industry.
文摘Knowing the influence of the size of datasets for regression models can help in improving the accuracy of a solar power forecast and make the most out of renewable energy systems.This research explores the influence of dataset size on the accuracy and reliability of regression models for solar power prediction,contributing to better forecasting methods.The study analyzes data from two solar panels,aSiMicro03036 and aSiTandem72-46,over 7,14,17,21,28,and 38 days,with each dataset comprising five independent and one dependent parameter,and split 80–20 for training and testing.Results indicate that Random Forest consistently outperforms other models,achieving the highest correlation coefficient of 0.9822 and the lowest Mean Absolute Error(MAE)of 2.0544 on the aSiTandem72-46 panel with 21 days of data.For the aSiMicro03036 panel,the best MAE of 4.2978 was reached using the k-Nearest Neighbor(k-NN)algorithm,which was set up as instance-based k-Nearest neighbors(IBk)in Weka after being trained on 17 days of data.Regression performance for most models(excluding IBk)stabilizes at 14 days or more.Compared to the 7-day dataset,increasing to 21 days reduced the MAE by around 20%and improved correlation coefficients by around 2.1%,highlighting the value of moderate dataset expansion.These findings suggest that datasets spanning 17 to 21 days,with 80%used for training,can significantly enhance the predictive accuracy of solar power generation models.
基金support from the National Key Scientific and Technological Infrastructure project “Earth System Science Numerical Simulator Facility” (Earth Lab)sponsored by the National Natural Science Foundation of China (Grant Nos. 42175132, 92044303, and 42205119)+2 种基金the National Key R&D Program (Grant Nos. 2020YFA0607802 and 2022YFC3703003)the CAS Information Technology Program (Grant No. CAS-WX2021SF-0107-02)the fellowship of China Postdoctoral Science Foundation (Grant No. 2022M723093)
文摘Scientific knowledge on the chemical compositions of fine particulate matter(PM_(2.5)) is essential for properly assessing its health and climate effects,and for decisionmakers to develop efficient mitigation strategies.A high-resolution PM_(2.5) chemical composition dataset(CAQRA-aerosol)is developed in this study,which provides hourly maps of organic carbon,black carbon,ammonium,nitrate,and sulfate in China from 2013 to 2020 with a horizontal resolution of 15 km.This paper describes the method,access,and validation results of this dataset.It shows that CAQRA-aerosol has good consistency with observations and achieves higher or comparable accuracy with previous PM_(2.5) composition datasets.Based on CAQRA-aerosol,spatiotemporal changes of different PM_(2.5) compositions were investigated from a national viewpoint,which emphasizes different changes of nitrate from other compositions.The estimated annual rate of population-weighted concentrations of nitrate is 0.23μg m^(−3)yr^(−1) from 2015 to 2020,compared with−0.19 to−1.1μg m^(−3)yr^(−1) for other compositions.The whole dataset is freely available from the China Air Pollution Data Center(https://doi.org/10.12423/capdb_PKU.2023.DA).
基金supported by the National Natural Science Foundation of China(42030102,42371321).
文摘Accurate fine-grained geospatial scene classification using remote sensing imagery is essential for a wide range of applications.However,existing approaches often rely on manually zooming remote sensing images at different scales to create typical scene samples.This approach fails to adequately support the fixed-resolution image interpretation requirements in real-world scenarios.To address this limitation,we introduce the million-scale fine-grained geospatial scene classification dataset(MEET),which contains over 1.03 million zoom-free remote sensing scene samples,manually annotated into 80 fine-grained categories.In MEET,each scene sample follows a scene-in-scene layout,where the central scene serves as the reference,and auxiliary scenes provide crucial spatial context for fine-grained classification.Moreover,to tackle the emerging challenge of scene-in-scene classification,we present the context-aware transformer(CAT),a model specifically designed for this task,which adaptively fuses spatial context to accurately classify the scene samples.CAT adaptively fuses spatial context to accurately classify the scene samples by learning attentional features that capture the relationships between the center and auxiliary scenes.Based on MEET,we establish a comprehensive benchmark for fine-grained geospatial scene classification,evaluating CAT against 11 competitive baselines.The results demonstrate that CAT significantly outperforms these baselines,achieving a 1.88%higher balanced accuracy(BA)with the Swin-Large backbone,and a notable 7.87%improvement with the Swin-Huge backbone.Further experiments validate the effectiveness of each module in CAT and show the practical applicability of CAT in the urban functional zone mapping.The source code and dataset will be publicly available at https://jerrywyn.github.io/project/MEET.html.
基金supported by the National Natural Science Foundation of China(22379157)CAS Project for Young Scientists in Basic Research(YSBR-102)+2 种基金Institute of Coal Chemistry,Chinese Academy of Sciences(SCJC-XCL-2023-13,SCJCXCL-2023-10)Talent Projects for Outstanding Doctoral Students to Work in Shanxi Province(E3SWR4791Z)Fundamental Research Program of Shanxi Province(202403021222485).
文摘This data set collects,compares and contrasts the capacities and structures of a series of hard carbon materials,and then searches for correlations between structure and electrochemical performance.The capacity data of the hard carbons were obtained by charge/discharge tests and the materials were characterized by XRD,gas adsorption,true density tests and SAXS.In particular,the fitting of SAXS gave a series of structural parameters which showed good characterization.The related test details are given with the structural data of the hard carbons and the electrochemical performance of the sodium-ion batteries.
文摘The increasing adoption of Industrial Internet of Things(IIoT)systems in smart manufacturing is leading to raise cyberattack numbers and pressing the requirement for intrusion detection systems(IDS)to be effective.However,existing datasets for IDS training often lack relevance to modern IIoT environments,limiting their applicability for research and development.To address the latter gap,this paper introduces the HiTar-2024 dataset specifically designed for IIoT systems.As a consequence,that can be used by an IDS to detect imminent threats.Likewise,HiTar-2024 was generated using the AREZZO simulator,which replicates realistic smart manufacturing scenarios.The generated dataset includes five distinct classes:Normal,Probing,Remote to Local(R2L),User to Root(U2R),and Denial of Service(DoS).Furthermore,comprehensive experiments with popular Machine Learning(ML)models using various classifiers,including BayesNet,Logistic,IBK,Multiclass,PART,and J48 demonstrate high accuracy,precision,recall,and F1-scores,exceeding 0.99 across all ML metrics.The latter result is reached thanks to the rigorous applied process to achieve this quite good result,including data pre-processing,features extraction,fixing the class imbalance problem,and using a test option for model robustness.This comprehensive approach emphasizes meticulous dataset construction through a complete dataset generation process,a careful labelling algorithm,and a sophisticated evaluation method,providing valuable insights to reinforce IIoT system security.Finally,the HiTar-2024 dataset is compared with other similar datasets in the literature,considering several factors such as data format,feature extraction tools,number of features,attack categories,number of instances,and ML metrics.
文摘This paper introduces a robust Distributed Denial-of-Service attack detection framework tailored for Software-Defined Networking based Internet of Things environments,built upon a novel,syntheticmulti-vector dataset generated in a Mininet-Ryu testbed using real-time flow-based labeling.The proposed model is based on the XGBoost algorithm,optimized with Principal Component Analysis for dimensionality reduction,utilizing lightweight flowlevel features extracted from Open Flow statistics to classify attacks across critical IoT protocols including TCP,UDP,HTTP,MQTT,and CoAP.The model employs lightweight flow-level features extracted from Open Flow statistics to ensure low computational overhead and fast processing.Performance was rigorously evaluated using key metrics,including Accuracy,Precision,Recall,F1-Score,False Alarm Rate,AUC-ROC,and Detection Time.Experimental results demonstrate the model’s high performance,achieving an accuracy of 98.93%and a low FAR of 0.86%,with a rapid median detection time of 1.02 s.This efficiency validates its superiority in meeting critical Key Performance Indicators,such as Latency and high Throughput,necessary for time-sensitive SDN-IoT systems.Furthermore,the model’s robustness and statistically significant outperformance against baseline models such as Random Forest,k-Nearest Neighbors,and Gradient Boosting Machine,validating through statistical tests using Wilcoxon signed-rank test and confirmed via successful deployment in a real SDN testbed for live traffic detection and mitigation.
基金supported by the Natural Science Basic Research Program of Shaanxi(Program No.2024JC-YBMS-026).
文摘When dealing with imbalanced datasets,the traditional support vectormachine(SVM)tends to produce a classification hyperplane that is biased towards the majority class,which exhibits poor robustness.This paper proposes a high-performance classification algorithm specifically designed for imbalanced datasets.The proposed method first uses a biased second-order cone programming support vectormachine(B-SOCP-SVM)to identify the support vectors(SVs)and non-support vectors(NSVs)in the imbalanced data.Then,it applies the synthetic minority over-sampling technique(SV-SMOTE)to oversample the support vectors of the minority class and uses the random under-sampling technique(NSV-RUS)multiple times to undersample the non-support vectors of the majority class.Combining the above-obtained minority class data set withmultiple majority class datasets can obtainmultiple new balanced data sets.Finally,SOCP-SVM is used to classify each data set,and the final result is obtained through the integrated algorithm.Experimental results demonstrate that the proposed method performs excellently on imbalanced datasets.
基金funded by A’Sharqiyah University,Sultanate of Oman,under Research Project grant number(BFP/RGP/ICT/22/490).
文摘Detecting faces under occlusion remains a significant challenge in computer vision due to variations caused by masks,sunglasses,and other obstructions.Addressing this issue is crucial for applications such as surveillance,biometric authentication,and human-computer interaction.This paper provides a comprehensive review of face detection techniques developed to handle occluded faces.Studies are categorized into four main approaches:feature-based,machine learning-based,deep learning-based,and hybrid methods.We analyzed state-of-the-art studies within each category,examining their methodologies,strengths,and limitations based on widely used benchmark datasets,highlighting their adaptability to partial and severe occlusions.The review also identifies key challenges,including dataset diversity,model generalization,and computational efficiency.Our findings reveal that deep learning methods dominate recent studies,benefiting from their ability to extract hierarchical features and handle complex occlusion patterns.More recently,researchers have increasingly explored Transformer-based architectures,such as Vision Transformer(ViT)and Swin Transformer,to further improve detection robustness under challenging occlusion scenarios.In addition,hybrid approaches,which aim to combine traditional andmodern techniques,are emerging as a promising direction for improving robustness.This review provides valuable insights for researchers aiming to develop more robust face detection systems and for practitioners seeking to deploy reliable solutions in real-world,occlusionprone environments.Further improvements and the proposal of broader datasets are required to developmore scalable,robust,and efficient models that can handle complex occlusions in real-world scenarios.
文摘Climate change significantly affects environment,ecosystems,communities,and economies.These impacts often result in quick and gradual changes in water resources,environmental conditions,and weather patterns.A geographical study was conducted in Arizona State,USA,to examine monthly precipi-tation concentration rates over time.This analysis used a high-resolution 0.50×0.50 grid for monthly precip-itation data from 1961 to 2022,Provided by the Climatic Research Unit.The study aimed to analyze climatic changes affected the first and last five years of each decade,as well as the entire decade,during the specified period.GIS was used to meet the objectives of this study.Arizona experienced 51–568 mm,67–560 mm,63–622 mm,and 52–590 mm of rainfall in the sixth,seventh,eighth,and ninth decades of the second millennium,respectively.Both the first and second five year periods of each decade showed accept-able rainfall amounts despite fluctuations.However,rainfall decreased in the first and second decades of the third millennium.and in the first two years of the third decade.Rainfall amounts dropped to 42–472 mm,55–469 mm,and 74–498 mm,respectively,indicating a downward trend in precipitation.The central part of the state received the highest rainfall,while the eastern and western regions(spanning north to south)had significantly less.Over the decades of the third millennium,the average annual rainfall every five years was relatively low,showing a declining trend due to severe climate changes,generally ranging between 35 mm and 498 mm.The central regions consistently received more rainfall than the eastern and western outskirts.Arizona is currently experiencing a decrease in rainfall due to climate change,a situation that could deterio-rate further.This highlights the need to optimize the use of existing rainfall and explore alternative water sources.
文摘Standardized datasets are foundational to healthcare informatization by enhancing data quality and unleashing the value of data elements.Using bibliometrics and content analysis,this study examines China's healthcare dataset standards from 2011 to 2025.It analyzes their evolution across types,applications,institutions,and themes,highlighting key achievements including substantial growth in quantity,optimized typology,expansion into innovative application scenarios such as health decision support,and broadened institutional involvement.The study also identifies critical challenges,including imbalanced development,insufficient quality control,and a lack of essential metadata—such as authoritative data element mappings and privacy annotations—which hampers the delivery of intelligent services.To address these challenges,the study proposes a multi-faceted strategy focused on optimizing the standard system's architecture,enhancing quality and implementation,and advancing both data governance—through authoritative tracing and privacy protection—and intelligent service provision.These strategies aim to promote the application of dataset standards,thereby fostering and securing the development of new productive forces in healthcare.
基金support from the National Natural Science Foundation of China(Nos.U24B2034,U2139204)the China Petroleum Science and Technology Innovation Fund(2021DQ02-0501)the Science and Technology Support Project of Langfang(2024011073).
文摘Lithology identification is a critical aspect of geoenergy exploration,including geothermal energy development,gas hydrate extraction,and gas storage.In recent years,artificial intelligence techniques based on drill core images have made significant strides in lithology identification,achieving high accuracy.However,the current demand for advanced lithology identification models remains unmet due to the lack of high-quality drill core image datasets.This study successfully constructs and publicly releases the first open-source Drill Core Image Dataset(DCID),addressing the need for large-scale,high-quality datasets in lithology characterization tasks within geological engineering and establishing a standard dataset for model evaluation.DCID consists of 35 lithology categories and a total of 98,000 high-resolution images(512×512 pixels),making it the most comprehensive drill core image dataset in terms of lithology categories,image quantity,and resolution.This study also provides lithology identification accuracy benchmarks for popular convolutional neural networks(CNNs)such as VGG,ResNet,DenseNet,MobileNet,as well as for the Vision Transformer(ViT)and MLP-Mixer,based on DCID.Additionally,the sensitivity of model performance to various parameters and image resolution is evaluated.In response to real-world challenges,we propose a real-world data augmentation(RWDA)method,leveraging slightly defective images from DCID to enhance model robustness.The study also explores the impact of real-world lighting conditions on the performance of lithology identification models.Finally,we demonstrate how to rapidly evaluate model performance across multiple dimensions using low-resolution datasets,advancing the application and development of new lithology identification models for geoenergy exploration.
基金supported by the National Natural Science Foundation of China(82270310)Noncommunicable Chronic Diseases-National Science and Technology Major Project(No.2024ZD0527000 and No.2024ZD0527005)+2 种基金Jiangsu Provincial key research and development program(BE2023662)General project of Nanjing Health Commission(YKK22166).
文摘Background:The life-course management of children with tetralogy of Fallot(TOF)has focused on demonstrating brain structural alterations,developmental trajectories,and cognition-related changes that unfold over time.Methods:We introduce an magnetic resonance imaging(MRI)dataset comprising TOF children who underwent brain MRI scanning and cross-sectional neurocognitive follow-up.The dataset includes brain three-dimensional T1-weighted imaging(3D-T1WI),three-dimensional T2-weighted imaging(3D-T2WI),and neurodevelopmental evaluations using the Wechsler Preschool and Primary Scale of Intelligence–Fourth Edition(WPPSI-IV).Results:Thirty-one children with TOF(age range:4–33 months;18 males)were recruited and completed corrective surgery at the Children’s Hospital of Nanjing Medical University,Nanjing,China.Aiming to promote the neurodevelopmental outcomes in children with TOF,we have meticulously curated a comprehensive dataset designed to dissect the complex interplay among risk factors,neuroimaging findings,and adverse neurodevelopmental outcomes.Conclusion:This article aims to introduce our open-source dataset on neurodevelopment in children with TOF,which covers the data types,data acquisition and processing methods,the procedure for accessing the data,and related publications.
文摘The aim of this article is to explore potential directions for the development of artificial intelligence(AI).It points out that,while current AI can handle the statistical properties of complex systems,it has difficulty effectively processing and fully representing their spatiotemporal complexity patterns.The article also discusses a potential path of AI development in the engineering domain.Based on the existing understanding of the principles of multilevel com-plexity,this article suggests that consistency among the logical structures of datasets,AI models,model-building software,and hardware will be an important AI development direction and is worthy of careful consideration.
基金supported in part by NSFC under Grant Nos.62402379,U22A2029 and U24A20237.
文摘The unprecedented scale of large models,such as large language models(LLMs)and text-to-image diffusion models,has raised critical concerns about the unauthorized use of copyrighted data during model training.These concerns have spurred a growing demand for dataset copyright auditing techniques,which aim to detect and verify potential infringements in the training data of commercial AI systems.This paper presents a survey of existing auditing solutions,categorizing them across key dimensions:data modality,model training stage,data overlap scenarios,and model access levels.We highlight major trends,including the prevalence of black-box auditing methods and the emphasis on fine-tuning rather than pre-training.Through an in-depth analysis of 12 representative works,we extract four key observations that reveal the limitations of current methods.Furthermore,we identify three open challenges and propose future directions for robust,multimodal,and scalable auditing solutions.Our findings underscore the urgent need to establish standardized benchmarks and develop auditing frameworks that are resilient to low watermark densities and applicable in diverse deployment settings.
