检索结果-内蒙古大学图书馆

38th Conference on Neural Information Processing Systems, NeurIPS 2024

作者： Shi, Zhenning Zheng, Haoshuai Xu, Chen Dong, Changsheng Pan, Bin Xie, Xueshuo He, Along Li, Tao Fu, Huazhu College of Computer Science Nankai University China School of Statistics and Data Science Nankai University China Haihe Lab of ITAI China Institute of High Performance Computing A*STAR Singapore

Recently, research on denoising diffusion models has expanded its application to the field of image restoration. Traditional diffusion-based image restoration methods utilize degraded images as conditional input to effectively guide the reverse generation process, without modifying the original denoising diffusion process. However, since the degraded images already include low-frequency information, starting from Gaussian white noise will result in increased sampling steps. We propose Resfusion, a general framework that incorporates the residual term into the diffusion forward process, starting the reverse process directly from the noisy degraded images. The form of our inference process is consistent with the DDPM. We introduced a weighted residual noise, named resnoise, as the prediction target and explicitly provide the quantitative relationship between the residual term and the noise term in resnoise. By leveraging a smooth equivalence transformation, Resfusion determine the optimal acceleration step and maintains the integrity of existing noise schedules, unifying the training and inference processes. The experimental results demonstrate that Resfusion exhibits competitive performance on ISTD dataset, LOL dataset and Raindrop dataset with only five sampling steps. Furthermore, Resfusion can be easily applied to image generation and emerges with strong versatility. Our code and model are available at https://***/nkicsl/Resfusion. © 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

Iterative methods via locally evolving set process 24

Iterative methods via locally evolving set process

引用

Proceedings of the 38th International Conference on Neural Information Processing Systems

作者： Baojian Zhou Yifan Sun Reza Babanezhad Harikandeh Xingzhi Guo Deqing Yang Yanghua Xiao The School of Data Science Fudan University and Shanghai Key Laboratory of Data Science School of Computer Science Fudan University Department of Computer Science Stony Brook University Samsung SAIT AI Lab. Shanghai Key Laboratory of Data Science School of Computer Science Fudan University

ISBN: (纸本)9798331314385

Given the damping factor α and precision tolerance ε, Andersen et al. [2] introduced Approximate Personalized PageRank (APPR), the de facto local method for approximating the PPR vector, with runtime bounded by Θ(1/(αε)) independent of the graph size. Recently, Fountoulakis & Yang [12] asked whether faster local algorithms could be developed using Õ(1/(/√αε)) operations. By noticing that APPR is a local variant of Gauss-Seidel, this paper explores the question of whether standard iterative solvers can be effectively localized. We propose to use the locally evolving set process, a novel framework to characterize the algorithm locality, and demonstrate that many standard solvers can be effectively localized. Let $\overline{\operatorname{vol}}{ (\mathcal S_t)}$ and $\overline{\gamma_t}$ be the running average of volume and the residual ratio of active nodes St during the process. We show $\overline{\operatorname{vol}}{ (\mathcal S_t)}/\overline{\gamma_t} \leq 1/\epsilon$ and prove APPR admits a new runtime bound $\tilde{\mathcal{O}}(\overline{\operatorname{vol}}(\mathcal S_t)/(\alpha\overline{\gamma_t}))$ mirroring the actual performance. Furthermore, when the geometric mean of residual reduction is Θ(√α), then there exists c ∈ (0, 2) such that the local Chebyshev method has runtime $\tilde{\mathcal{O}}(\overline{\operatorname{vol}}(\mathcal{S_t})/(\sqrt{\alpha}(2-c)))$ without the monotonicity assumption. Numerical results confirm the efficiency of this novel framework and show up to a hundredfold speedup over corresponding standard solvers on real-world graphs.

关键词：

来源：评论

学校读者我要写书评

暂无评论

An Intrusion Detection Scheme Based on Sequential Tri-Way Decisions of Multiple Granularity for Intelligent Connected Vehicles

An Intrusion Detection Scheme Based on Sequential Tri-Way De...

引用

International Conference on Artificial Intelligence and Autonomous Transportation, AIAT 2024

作者： Li, Yongzhong Wang, Qi Li, Yuheng Zhang, Shipeng School of Automotive Engineering Nantong Institute of Technology Nantong226010 China Internet of Things and Big Data Major of Macao Polytechnic University Macao China School of Computer Science Jiangsu University of Science and Technology Zhenjiang212003 China

ISBN: (纸本)9789819639762

With the rapid developing technology of the Internet of Things, intelligent connected vehicles are facing more and more serious cybersecurity threats and challenges. With the developing of cybersecurity technology for intelligent connected vehicles it has posed a comprehensive challenge to the existing cybersecurity technologies. Intrusion detection is an important research field of cybersecurity. An intrusion detection scheme based on sequential tri-way decisions of multiple granularities (S3WD-MG) is proposed for Intelligent Connected Vehicles. The S3WD-MG method obtains feature sets of multi-granularities through the auto-encoder of deep learning. This method combines multiple granularities and sequential tri-way decisions to make decisions by the most appropriate for network behavior through by the decision threshold. The experiment results on NSL-KDD show that S3WD-MG method does have a good performance in intrusion detection. Compared with other models, our proposed S3WD-MG method has a higher detection’s rate and the robustness of algorithm is stronger. © Beijing Paike Culture Commu. Co., Ltd. 2025.

关键词： Network intrusion

来源：评论

学校读者我要写书评

暂无评论

Predicting Protein-Protein Interactions Using Self-Attention-Based Deep Neural Networks and FastText Embeddings 15

Predicting Protein-Protein Interactions Using Self-Attention...

引用

15th International Conference on Computing Communication and Networking Technologies, ICCCNT 2024

作者： Oviya, I.R. Shanmukha Sravya, N. Raja, Kalpana Amrita Vishwa Vidyapeetham Amrita School Of Computing Department of Computer Science Engineering Chennai601103 India Yale University Section for Biomedical Informatics and Data Science School of Medicine New HavenCT United States

ISBN: (纸本)9798350370249

Predicting Protein-protein interactions are central to understanding the intricate mechanisms governing cellular functions, as proteins seldom act in isolation. While current machine learning and deep learning models focus on encoding protein sequences based on residue composition or physicochemical properties, they often overlook the complex relationships between amino acid segments within lengthy sequences. To address these limitations, we propose a novel hybrid model that combines a self-attention to enhance Deep neural network feature extraction with an with FastText Protein embeddings for predicting PPI. Utilizing transfer learning via the unsupervised FastText embedding technique to derive statistical representations of protein sequences, offers advantages over traditional protein encoding techniques in the context of analysing k-Mer sequences. Our model's performance is assessed through 5-fold cross-validation on datasets of *** and Human-Bacillus Anthracis protein interactions, resulting in accuracies of 92.28% and 94.08%, respectively. Compared to existing machine learning models, our model demonstrates superior performance on both datasets. © 2024 IEEE.

关键词： Adversarial machine learning

来源：评论

学校读者我要写书评

暂无评论

Locating What You Need: Towards Adapting Diffusion Models to OOD Concepts In-the-Wild 38

Locating What You Need: Towards Adapting Diffusion Models to...

引用

38th Conference on Neural Information Processing Systems, NeurIPS 2024

作者： Yang, Jianan Gao, Chenchao Xiao, Zhiqing Zhao, Junbo Wu, Sai Chen, Gang Wang, Haobo College of Computer Science and Technology Zhejiang University China Institute of Blockchain and Data Security China School of Software Technology Zhejiang University China International School of Information Science and Engineering Dalian University of Technology China

The recent large-scale text-to-image generative models have attained unprecedented performance, while people established adaptor modules like LoRA and DreamBooth to extend this performance to even more unseen concept tokens. However, we empirically find that this workflow often fails to accurately depict the out-of-distribution concepts. This failure is highly related to the low quality of training data. To resolve this, we present a framework called Controllable Adaptor Towards Out-of-Distribution Concepts (CATOD). Our framework follows the active learning paradigm which includes high-quality data accumulation and adaptor training, enabling a finer-grained enhancement of generative results. The aesthetics score and concept-matching score are two major factors that impact the quality of synthetic results. One key component of CATOD is the weighted scoring system that automatically balances between these two scores and we also offer comprehensive theoretical analysis for this point. Then, it determines how to select data and schedule the adaptor training based on this scoring system. The extensive results show that CATOD significantly outperforms the prior approaches with an 11.10 boost on the CLIP score and a 33.08% decrease on the CMMD metric. © 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

VDDA: An Effective Software Vulnerability Detection Model Based on Deep Learning and Attention Mechanism 26

VDDA: An Effective Software Vulnerability Detection Model Ba...

引用

26th International Conference on computer Supported Cooperative Work in Design, CSCWD 2023

作者： Chang, Jiaqi Ma, Zhujuan Cao, Binghao Zhu, Erzhou Anhui University School of Computer Science and Technology Hefei China Anhui Xinhua University School of Big Data and Artificial Intelligence Hefei China

ISBN: (纸本)9798350331684

Many of software vulnerability detection methods suffer from problems of dependent on expert experience, rough detection granularity, and incomplete syntax and semantics information on source codes. This paper proposes the VDDA, Vulnerability Detection based on Deep learning and Attention mechanism, an effective software vulnerability detection model based on deep learning and the attention mechanism. In the VDDA, deep learning technology is used to construct the underlying classifier to avoid the feature engineering of traditional machine learning techniques. The Joren slice tool combined with the code attribute graph (CPG) optimization is used to simplify the source code before it is fed to the Bidirectional Long Short-Term Memory (BLSTM) deep model. Meanwhile, the attention mechanism is employed to improve the efficiency and accuracy of vulnerability detection. Experiment results have demonstrated that the proposed VDDA model is more effective than the existing vulnerability detection methods. © 2023 IEEE.

关键词： Semantics

来源：评论

学校读者我要写书评

暂无评论

Learning to Unlearn for Bayesian Personalized Ranking via Influence Function

引用

Chinese Journal of Electronics 2025年第3期34卷 990-1001页

作者： Chen Jundong Zhang Honglei Li Haoxuan Li Yidong Key Laboratory of Big Data & Artificial Intelligence in Transportation (Beijing Jiaotong University) Ministry of Education Beijing 100044 China School of Cyberspace Science and Technology Beijing Jiaotong University Beijing 100044 China School of Computer Science and Technology Beijing Jiaotong University Beijing 100044 China Center for Data Science Peking University Beijing 100871 China

Learning recommender models from vast amounts of behavioral data has become a mainstream paradigm in recent information systems. Conversely, with privacy awareness grown, there has been increasing attention to the removal of sensitive or outlier data from well-trained recommendation models (known as recommendation unlearning). However, current unlearning methods primarily focus on fully/partially retraining the entire model. Despite considerable performance, it inevitably introduces significant efficiency bottlenecks, which is impractical for latency-sensitive streaming services. While recent efforts exploit efficient unlearning in point-wise recommender tasks, these approaches overlook the partial order relationships between items, resulting in suboptimal performance in both recommendation and unlearning capabilities. In light of this, we explore learning to unlearn for Bayesian personalized ranking via influence function, which relies on a pair-wise ranking loss to model user preferences and item characteristics, making unlearning more challenging than in point-wise settings. Specifically, we propose an influence function-guided unlearning framework tailored for pair-wise ranking models to efficiently perform unlearning requests, which involves unlearning partial order relationships while handling negative samples appropriately during the unlearning process. Besides, we prove that our proposed method can theoretically match the performance of retraining counterparts. Finally, we conduct extensive experiments to validate the effectiveness and efficiency of our model.

关键词： Recommendation unlearning Influence function Bayesian personalized ranking The right to be forgotten

来源：评论

学校读者我要写书评

暂无评论

Effective In-Context Learning for Named Entity Recognition

Effective In-Context Learning for Named Entity Recognition

引用

2024 IEEE International Conference on Bioinformatics and Biomedicine, BIBM 2024

作者： Zhao, Jin Guo, Qian Liang, Jiaqing Li, Zhixu Xiao, Yanghua Fudan University Shanghai Key Laboratory of Data Science School of Computer Science Shanghai China Fuyang Normal University College of Information Engineering Anhui China Renmin University of China School of Information Suzhou China

ISBN: (纸本)9798350386226

Named entity recognition aims to identify meaningful words or phrases from text. Traditional NER methods depend heavily on training datasets to learn entity features, which often leads to poor performance in recognizing low-resource entities. Large language models offer a promising solution in low-resource scenarios, but they struggle with entity boundary recognition and often suffer from hallucination issues due to their reliance on general-purpose corpora. In this paper, we propose a novel approach that enhances LLMs' performance in NER by integrating a boundary regularity extraction algorithm and an explanatory path extraction method. The boundary regularity algorithm leverages ChatGPT's summarization capabilities to articulate entity boundaries in natural language, improving boundary detection accuracy. The explanatory path method constructs logical reasoning steps from text to entity recognition, guiding LLMs to generate more accurate predictions and mitigating hallucination problems. Extensive experiments on three public datasets demonstrate that our method outperforms current mainstream LLMs in accurately recognizing entities from unstructured text. These findings validate the effectiveness of our approach and its potential to address key challenges in NER. © 2024 IEEE.

关键词： Large language models Low-resource Named entity recognition

来源：评论

学校读者我要写书评

暂无评论

PREVALENCE OF TRESILLO RHYTHM IN CONTEMPORARY POPULAR MUSIC 21

PREVALENCE OF TRESILLO RHYTHM IN CONTEMPORARY POPULAR MUSIC

引用

21st Sound and Music Computing Conference, SMC 2024

作者： Jajoria, Pushkar Mäder, Aurel Ruben Krenn, Florian McDermott, James Data Science Institute School of Computer Science University of Galway Galway Ireland Digital and Cognitive Musicology Lab École Polytechnique Fédérale de Lausanne Lausanne Switzerland

ISBN: (纸本)9789893520758

This paper presents an investigation into the prevalence and significance of the Tresillo rhythm in contemporary popular music. Initially, the Tresillo rhythm is defined and formalized within a computational framework. Subsequently, employing mathematical representations, its presence is traced within the US Billboard Top 20 Charts spanning the last two decades. Detection and quantification of the Tresillo rhythm within musical compositions are achieved through the calculation of similarities between established formalizations of the rhythm and individual songs. The resulting similarity metrics provide insights into the degree of resemblance between a given pop song's rhythm and the predefined Tresillo pattern. To ensure the robustness of the findings, two distinct formalizations of the Tresillo rhythm are constructed, and multiple methodologies for computing rhythm similarities are systematically assessed and compared. Leveraging these similarity measures, an empirical examination of Tresillo rhythm usage in the US Billboard Top 20 Charts from 1999 to 2019 is conducted. Finally, potential explanations for observed trends are discussed and analyzed. © 2024. This is an open-access article distributed under the terms of the Creative Commons Attribution 3.0 Unported License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original.

关键词： Music

来源：评论

学校读者我要写书评

暂无评论

A Dual-Defense Self-balancing Framework Against Bilateral Model Attacks in Federated Learning 24th

A Dual-Defense Self-balancing Framework Against Bilateral Mo...

引用

24th International Conference on Algorithms and Architectures for Parallel Processing, ICA3PP 2024

作者： Wu, Xiang Yao, Aiting Pal, Shantanu Jiang, Frank Li, Xuejun Xu, Jia Dong, Chengzu Chen, Xuefei Zhang, Xiuyi Liu, Xiao Anhui Provincial International Joint Research Center for Advanced Technology in Medical Imaging School of Computer Science and Technology Anhui University Anhui China Deakin University Melbourne Australia School of Data Science Lingnan University Hong Kong

ISBN: (纸本)9789819615247

With the rapid expansion of Artificial Intelligence (AI) services, smart devices generate a large amount of user data at the edge network, which urgently needs to be protected while effectively extracting information. Federated learning (FL) is an important technology for handling dispersed data and strict privacy requirements in this context. However, the security threats caused by model inversion attacks and poisoning attacks can affect the mutual trust between the client and server. Yet, for these two types of attacks, the existing defense mechanisms are contradictory in terms of whether the model parameters are publicly disclosed. In addition, the data distribution of the clients is imbalanced which will increase the bias of model, reducing its practicality. To address this issue, this study proposes a dual defense self-balanced federated learning (DDSFL) framework, aiming to introduce a novel lightweight defense mechanism during the model parameter aggregation stage, combating these two types of attacks simultaneously by applying differential privacy and adjusting learning rates. In addition, this method also integrates a middleware-based reordering algorithm to enhance the robustness of the framework. Experimental results show that DDSFL effectively improves the ability to resist imbalanced data, forged data, and malicious behavior, significantly enhancing the generalization performance and security of the FL system. © The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2025.

关键词： Federated learning

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：