检索结果-内蒙古大学图书馆

A federated learning scheme meets dynamic differential privacy

CAAI Transactions on Intelligence Technology 2023年第3期8卷 1087-1100页

作者： Shengnan Guo Xibin Wang Shigong Long Hai Liu Liu Hai Toong Hai Sam State Key Laboratory of Public Big Data College of Computer Science and TechnologyGuizhou UniversityGuiyangChina School of Big Data Key Laboratory of Electric Power Big Data of Guizhou ProvinceGuizhou Institute of TechnologyGuiyangChina School of Information Guizhou University of Finance and EconomicsGuiyangChina Faculty of Business and Communication INTI International UniversityNilaiMalaysia

Federated learning is a widely used distributed learning approach in recent years,however,despite model training from collecting data become to gathering parameters,privacy violations may occur when publishing and sharing models.A dynamic approach is pro-posed to add Gaussian noise more effectively and apply differential privacy to federal deep ***,it is abandoning the traditional way of equally distributing the privacy budget e and adjusting the privacy budget to accommodate gradient descent federation learning dynamically,where the parameters depend on computation derived to avoid the impact on the algorithm that hyperparameters are created *** also incorporates adaptive threshold cropping to control the sensitivity,and finally,moments accountant is used to counting the∈consumed on the privacy‐preserving,and learning is stopped only if the∈_(total)by clients setting is reached,this allows the privacy budget to be adequately explored for model *** experimental results on real datasets show that the method training has almost the same effect as the model learning of non‐privacy,which is significantly better than the differential privacy method used by TensorFlow.

关键词： data privacy machine learning security of data

来源：评论

学校读者我要写书评

暂无评论

ON STOCHASTIC CONTROL PROBLEMS WITH HIGHER-ORDER MOMENTS

引用

SIAM Journal on Control and Optimization 2025年第3期63卷 1560-1589页

作者： Wang, Yike Liu, Jingzhen Bensoussan, Alain Cedric Yiu, Ka-Fai Wei, Jiaqin School of Finance Chongqing Technology and Business University Chongqing China China Institute for Actuarial Science Central University of Finance and Economics Beijing China SNaveen Jindal School of Management University of Texas at Dallas DallasTX75080 United States Department of Applied Mathematics Hong Kong Polytechnic University Hong Kong Key Laboratory of Advanced Theory and Application in Statistics and Data Science-MOE School of Statistics East China Normal University Shanghai China

In this paper, we focus on a class of time-inconsistent stochastic control problems, where the objective function includes the mean and several higher-order central moments of the terminal value of state. To tackle the time-inconsistency, we seek both the closed-loop and the open-loop Nash equilibrium controls as time-consistent solutions. We establish a partial differential equation (PDE) system for deriving a closed-loop Nash equilibrium control, which does not include the equilibrium value function and is different from the extended Hamilton-Jacobi-Bellman (HJB) equations as in Bj\" ork, Khapko, and Murgoci [finance Stoch., 21 (2017), pp. 331-360]. We show that our PDE system is equivalent to the extended HJB equations that seem difficult to solve for our higher-order moment problems. In deriving an open-loop Nash equilibrium control, due to the nonseparable higher-order moments in the objective function, we make some moment estimates in addition to the standard perturbation argument for developing a maximum principle. Then, the problem is reduced to solving a flow of forward-backward stochastic differential equations. In particular, we investigate linear controlled dynamics and some objective functions affine in the mean. The closed-loop and the open-loop Nash equilibrium controls are identical, which are independent of the state value, the random path, and the preference on the odd-order central moments. By sending the highest order of central moments to infinity, we obtain the time-consistent solutions to some control problems whose objective functions include some penalty functions for deviation. © 2025 Society for Industrial and Applied Mathematics.

关键词： Stochastic control systems

来源：评论

学校读者我要写书评

暂无评论

Deep nonparametric quantile regression under covariate shift

The Journal of Machine Learning Research

引用

The Journal of Machine Learning Research 2024年第1期25卷 18851-18900页

作者： Xingdong Feng Xin He Yuling Jiao Lican Kang Caixing Wang School of Statistics and Data Science & Institute of Data Science and Statistics Shanghai University of Finance and Economics Shanghai China School of Statistics and Data Science Shanghai University of Finance and Economics Shanghai China School of Artificial Intelligence Hubei Key Laboratory of Computational Science Wuhan University Wuhan China Institute for Math and AI School of Mathematics and Statistics Wuhan University Wuhan China

This work focuses on addressing the challenges posed by covariate shift in nonparametric quantile regression using deep neural networks. We propose a two-stage pretraining reweighted method that leverages importance weighting to mitigate the effects of distribution shift. In the first stage, density ratios are estimated with a neural network by minimizing least squares. In the second stage, a deep neural network estimator is obtained using pre-training weights. Theoretical analysis is provided, offering non-asymptotic error bounds for the unweighted, reweighted, and pre-training reweighted estimators. We consider scenarios with both bounded and unbounded density ratios. Notably, we employ a novel proof technique to bound the generalization error, characterized by the size and weights bound of ReLU neural networks. This enables us to establish fast rates of convergence under the adaptive self-calibration condition, distinguishing our approach from those relying on local Rademacher complexity techniques. Additionally, we derive the approximation error with weight bounds for ReLU neural networks approximating the Hölder class. Our theoretical findings provide valuable insights for the pre-training process and highlight the efficacy of reweighted techniques. Numerical experiments are conducted to further validate the theoretical findings and demonstrate the effectiveness of our proposed method.

关键词： deep neural networks distribution shift non-asymptotic error bounds robust estimation

来源：评论

学校读者我要写书评

暂无评论

On the Noise Robustness of In-Context Learning for Text Generation 38

On the Noise Robustness of In-Context Learning for Text Gene...

引用

38th Conference on Neural Information Processing Systems, NeurIPS 2024

作者： Gao, Hongfu Zhang, Feipeng Jiang, Wenyu Shu, Jun Zheng, Feng Wei, Hongxin Department of Statistics and Data Science Southern University of Science and Technology China School of Economics and Finance Xi'an Jiaotong University China National Key Laboratory for Novel Software Technology Nanjing University China School of Mathematics and Statistics Xi'an Jiaotong University China Department of Computer Science and Engineering Southern University of Science and Technology China

Large language models (LLMs) have shown impressive performance on downstream tasks by in-context learning (ICL), which heavily relies on the quality of demonstrations selected from a large set of annotated examples. Recent works claim that in-context learning is robust to noisy demonstrations in text classification. In this work, we show that, on text generation tasks, noisy annotations significantly hurt the performance of in-context learning. To circumvent the issue, we propose a simple and effective approach called Local Perplexity Ranking (LPR), which replaces the "noisy" candidates with their nearest neighbors that are more likely to be clean. Our method is motivated by analyzing the perplexity deviation caused by noisy labels and decomposing perplexity into inherent perplexity and matching perplexity. Our key idea behind LPR is thus to decouple the matching perplexity by performing the ranking among the neighbors in semantic space. Our approach can prevent the selected demonstrations from including mismatched input-label pairs while preserving the effectiveness of the original selection methods. Extensive experiments demonstrate the effectiveness of LPR, improving the EM score by up to 18.75 on common benchmarks with noisy annotations. Our code is available at https://***/ml-stat-Sustech/Local-Perplexity-Ranking. © 2024 Neural information processing systems foundation. All rights reserved.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Rules still work for Open Information Extraction

arXiv

引用

arXiv 2024年

作者： Hua, Jialin Luo, Liangqing Ping, Weiying Liao, Yan Tao, Chunhai Lu, Xuewen Key Laboratory of Data Science in Finance and Economics School of Statistics and Data Science Jiangxi University of Finance and Economics Nanchang China Department of Mathematics and Statistics University of Calgary Calgary Canada

Open information extraction (OIE) aims to extract surface relations and their corresponding arguments from natural language text, irrespective of domain. This paper presents an innovative OIE model, APRCOIE, tailored for Chinese text. Diverging from previous models, our model generates extraction patterns autonomously. The model defines a new pattern form for Chinese OIE and proposes an automated pattern generation methodology. In that way, the model can handle a wide array of complex and diverse Chinese grammatical phenomena. We design a preliminary filter based on tensor computing to conduct the extraction procedure efficiently. To train the model, we manually annotated a large-scale Chinese OIE dataset. In the comparative evaluation, we demonstrate that APRCOIE outperforms state-of-the-art Chinese OIE models and significantly expands the boundaries of achievable OIE performance. The code of APRCOIE and the annotated dataset are released on GitHub (https://***/jialin666/APRCOIE_v1) © 2024, CC BY.

关键词： Tensors

来源：评论

学校读者我要写书评

暂无评论

Controlled-source electromagnetic noise attenuation via a deep convolutional neural network and high-quality sounding curve screening mechanism

引用

Geophysics 2025年第3期90卷 WA125-WA140页

作者： Liu, Yecheng Li, Diquan Li, Jin Zhang, Xian Central South University Monitoring Ministry of Education Key Laboratory of Metallogenic Prediction of Nonferrous Metals and Geological Environment Changsha China Hunan Provincial Key Laboratory of Non-ferrous Resources and Geological Hazard Detection Changsha China Central South University School of Geoscience and Info-physics Changsha China Hunan Normal University College of Information Science and Engineering Hunan Provincial Key Laboratory of Intelligent Computing and Language Information Processing Changsha China Hunan University of Finance and Economics School of Information Technology and Management Hunan Provincial Key Laboratory of Finance & Economics Big Data Science and Technology Changsha China

Strong noise is one of the biggest challenges in controlled-source electromagnetic (CSEM) exploration, which severely affects the quality of the recorded signal. We develop a novel and effective CSEM noise attenuation method that integrates a deep convolutional neural network (DCNN) with a high-quality sounding curve screening mechanism. This method first uses a DCNN to learn the large-scale noise characteristics of CSEM measured data and build a DCNN denoising model for attenuating the noise. Because some impulse noise remains after DCNN denoising, the high-quality sounding curve screening mechanism is innovatively applied to optimize the DCNN denoised data. The core of this method is the smoothness value (SV), which quantitatively evaluates the smoothness of the sounding curve (apparent resistivity curve). In effect, the mechanism segments the overall DCNN denoised data and calculates the corresponding sounding curve and its SV for each segment. The highest-quality sounding curve can be screened out by comparing the SVs of all curves as well as obtaining the highest-quality data segment present in the overall DCNN denoised data. Finally, our method is validated with synthetic and field data, demonstrating its feasibility and effectiveness in attenuating large-scale CSEM noise. The quality of CSEM data, which is heavily affected by noise, is significantly improved after processing by our method. © 2025 Society of Exploration Geophysicists. All rights reserved.

关键词： Deep neural networks

来源：评论

学校读者我要写书评

暂无评论

Lightweight Remote Sensing Target Detection Algorithm Based on Improved YOLOv5 16th

Lightweight Remote Sensing Target Detection Algorithm Based ...

引用

16th International Conference on Genetic and Evolutionary Computing, ICGEC 2024

作者： Wang, Chaomin Dong, Fang Liang, Sheng Yu, Wanxin Wu, Kaixian Yu, Long Wenfeng, Wang School of Information Engineering Nanchang Institute of Technology Nanchang330099 China Jiangxi Province Key Laboratory of Smart Water Conservancy Nanchang Institute of Technology Nanchang330099 China School of Statistics and Data Science Jiangxi University of Finance and Economics Nanchang330013 China

ISBN: (纸本)9789819615346

Aiming at the problems of large memory space occupied by remote sensing image detection models in complex scenes, high computational complexity, and detection speed that is difficult to meet the real-time target detection requirements, a model compression algorithm based on dynamic detection head and channel pruning is proposed. Based on YOLOv5s, the head of the original YOLOv5 is replaced by the DyHead detection head, which improves the detection performance of the model by combining and unifying the three attention mechanisms of scale, space and task. Secondly, the Focaler-IoU loss function is introduced to further improve the predictive regression ability of the model. Finally, the channel pruning algorithm is used to eliminate the redundant parameters in the network to further compress the parameters and computation of the model, and the accuracy of the pruned model is improved by model fine-tuning. The experimental results show that there is a large improvement on the DOTAV1.0 dataset, and the validity is further verified on the RSOD dataset. © The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2025.

关键词： Jurassic

来源：评论

学校读者我要写书评

暂无评论

Combining Experimental and Historical data for Policy Evaluation 41

Combining Experimental and Historical Data for Policy Evalua...

引用

41st International Conference on Machine Learning, ICML 2024

作者： Li, Ting Shi, Chengchun Wen, Qianglin Sui, Yang Qin, Yongli Lai, Chunbo Zhu, Hongtu School of Statistics and Management Shanghai University of Finance and Economics China Department of Statistics London School of Economics and Political Science United Kingdom Yunnan Key Laboratory of Statistical Modeling and Data Analysis Yunnan University China Didi Chuxing China Department of Biostatistics The University of North Carolina Chapel Hill United States

This paper studies policy evaluation with multiple data sources, especially in scenarios that involve one experimental dataset with two arms, complemented by a historical dataset generated under a single control arm. We propose novel data integration methods that linearly integrate base policy value estimators constructed based on the experimental and historical data, with weights optimized to minimize the mean square error (MSE) of the resulting combined estimator. We further apply the pessimistic principle to obtain more robust estimators, and extend these developments to sequential decision making. Theoretically, we establish non-asymptotic error bounds for the MSEs of our proposed estimators, and derive their oracle, efficiency and robustness properties across a broad spectrum of reward shift scenarios. Numerical experiments and real-data-based analyses from a ridesharing company demonstrate the superior performance of the proposed estimators. Copyright 2024 by the author(s)

关键词： data assimilation

来源：评论

学校读者我要写书评

暂无评论

CausalDiff: Causality-Inspired Disentanglement via Diffusion Model for Adversarial Defense 38

CausalDiff: Causality-Inspired Disentanglement via Diffusion...

引用

38th Conference on Neural Information Processing Systems, NeurIPS 2024

作者： Zhang, Mingkun Bi, Keping Chen, Wei Chen, Quanrun Guo, Jiafeng Cheng, Xueqi CAS Key Laboratory of AI Safety Institute of Computing Technology CAS China Key Laboratory of Network Data Science and Technology Institute of Computing Technology CAS China School of Statistics University of International Business and Economics China

Despite ongoing efforts to defend neural classifiers from adversarial attacks, they remain vulnerable, especially to unseen attacks. In contrast, humans are difficult to be cheated by subtle manipulations, since we make judgments only based on essential factors. Inspired by this observation, we attempt to model label generation with essential label-causative factors and incorporate label-non-causative factors to assist data generation. For an adversarial example, we aim to discriminate the perturbations as non-causative factors and make predictions only based on the label-causative factors. Concretely, we propose a casual diffusion model (CausalDiff) that adapts diffusion models for conditional data generation and disentangles the two types of casual factors by learning towards a novel casual information bottleneck objective. Empirically, CausalDiff has significantly outperformed state-of-the-art defense methods on various unseen attacks, achieving an average robustness of 86.39% (+4.01%) on CIFAR-10, 56.25% (+3.13%) on CIFAR-100, and 82.62% (+4.93%) on GTSRB (German Traffic Sign Recognition Benchmark). The code is available at https://***/CAS-AISafetyBasicResearchGroup/CausalDiff. © 2024 Neural information processing systems foundation. All rights reserved.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Improved Machine Learning Approach for Shanghai City's Secondhand Housing Price Prediction

Improved Machine Learning Approach for Shanghai City's Secon...

引用

2023 Congress in Computer science, Computer Engineering, and Applied Computing, CSCE 2023

作者： Hu, Guang Tang, Yue Yi, Rui School of Statistics and Information Shanghai University of International Business and Economics Shanghai China School of Computer Science Fudan University Shanghai China Shanghai Key Laboratory of Data Science Shanghai China

ISBN: (纸本)9798350327595

This paper presents an improved machine learning approach for prediction of second-hand housing prices in Shanghai. It firstly builds the random forest model and the XGboost model with Shanghai second-hand housing transaction data, and then analyses the challenges in the improvement of the models' prediction accuracy and generalization performance. In the light of that, it introduces the Lasso model for variable selection, and deletes three variables with insignificant regression coefficients, and then rebuilds the random forest and XGboost prediction models. The experimental results show that the prediction accuracy and the generalization performance of the models is well improved. © 2023 IEEE.

关键词： Lasso Machine Learning Prediction Random Forest XGboost

来源：评论

学校读者我要写书评

暂无评论

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案：

请选择收藏分类：

通借通还

建议与咨询 留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案： 新增检索档案 确定 取消

请选择收藏分类： 新增自定义分类 确定 取消

通借通还

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

请选择保存的检索档案：

请选择收藏分类：