检索结果-内蒙古大学图书馆

Learning-Based Directional Improvement Prediction for Dynamic Multiobjective Optimization

IEEE Transactions on Evolutionary Computation 2024年 1-1页

作者： Ye, Yulong Liu, Songbai Zhou, Junwei Lin, Qiuzhen Jiang, Min Tan, Kay Chen College of Computer Science and Software Engineering Shenzhen University Shenzhen China School of Computer Science and Artificial Intelligence Wuhan University of Technology Wuhan China School of Informatics Xiamen University Fujian China Department of Computing The Hong Kong Polytechnic University Hong Kong China

In recent years, dynamic multiobjective evolutionary algorithms (DMOEAs) using the prediction strategy have shown promising performance for solving dynamic multiobjective optimization problems (DMOPs), as they can predict environmental changing trends in advance. However, most of them follow a regular change pattern and thus their performance is compromised when solving DMOPs with irregular change patterns (e.g., nonlinear correlations). To alleviate this challenge, this article proposes a DMOEA with a learnable prediction for tackling DMOPs. Specifically, a neural network is designed to effectively capture diverse change patterns of the environment. Based on the change patterns learned, a directional improvement prediction (DIP) is developed to guide the evolutionary search toward promising directions in the decision space. In this way, a superior initial population with good convergence and diversity is predicted by DIP, which can be more effective for solving various DMOPs. Comprehensive empirical studies show that the proposed DIP is effective and the proposed algorithm has some advantages over five competitive DMOEAs when solving three commonly used benchmarks and one real-world problem. IEEE

关键词： Neural networks

来源：评论

学校读者我要写书评

暂无评论

MemSAM: Taming Segment Anything Model for Echocardiography Video Segmentation

MemSAM: Taming Segment Anything Model for Echocardiography V...

引用

Conference on computer Vision and Pattern Recognition (CVPR)

作者： Xiaolong Deng Huisi Wu Runhao Zeng Jing Qin College of Computer Science and Software Engineering Shenzhen University College of Mechatronics and Control Engineering Shenzhen University Centre for Smart Health School of Nursing The Hong Kong Polytechnic University

ISBN: (数字)9798350353006

ISBN: (纸本)9798350353013

We propose a novel echocardiographical video segmentation model by adapting SAM to medical videos to address some long-standing challenges in ultrasound video segmentation, including (1) massive speckle noise and artifacts, (2) extremely ambiguous boundaries, and (3) large variations of targeting objects across frames. The core technique of our model is a temporal-aware and noise-resilient prompting scheme. Specifically, we employ a space-time memory that contains both spatial and temporal information to prompt the segmentation of current frame, and thus we call the proposed model as MemSAM. In prompting, the memory carrying temporal cues sequentially prompt the video segmentation frame by frame. Meanwhile, as the memory prompt propagates high-level features, it avoids the issue of misidentification caused by mask propagation and improves representation consistency. To address the challenge of speckle noise, we further propose a memory reinforcement mechanism, which leverages predicted masks to improve the quality of the memory before storing it. We extensively evaluate our method on two public datasets and demonstrate state-of-the-art performance compared to existing models. Particularly, our model achieves comparable performance with fully supervised approaches with limited annotations. Codes are available at https://***/dengxl0520/MemSAM.

关键词： Image segmentation Adaptation models computer vision Ultrasonic imaging Echocardiography Annotations Noise

来源：评论

学校读者我要写书评

暂无评论

Semi-Asynchronous Federated Learning with Trajectory Prediction for Vehicular Edge Computing

Semi-Asynchronous Federated Learning with Trajectory Predict...

引用

International Workshop on Quality of Service

作者： Yuxuan Deng Xiuhua Li Chuan Sun Qilin Fan Xiaofei Wang Victor C. M. Leung School of Big Data & Software Engineering Chongqing University Chongqing China School of Computer Science and Engineering Nanyang Technological University Singapore TKLAN College of Intelligence & Computing Tianjin University Tianjin China Artificial Intelligence Research Institute Shenzhen MSU-BIT University Shenzhen China College of Computer Science & Software Engineering Shenzhen University Shenzhen China Department of Electrical & Computer Engineering The University of British Columbia Vancouver Canada

ISBN: (数字)9798350350128

ISBN: (纸本)9798350350135

Federated learning, as a distributed machine learning paradigm, offers promising solutions for vehicular edge computing (VEC) networks. However, federated learning in VEC with classification tasks still faces two key challenges: i) Delayed data labeling hampers supervised training; ii) Dynamic vehicle behavior complicates training scheduling and model uploads to edge servers. In this paper, we propose a semi-asynchronous federated learning algorithm for VEC. Specifically, it utilizes knowledge distillation to generate soft labels from raw data for supervised training, and estimates model training and uploading time through trajectory prediction. We further logically group vehicles based on the characteristics of their dynamic behavior. We then employ synchronous aggregation within groups and asynchronous aggregation between groups to optimize model performance while reducing latency. Finally, we conduct separate comparative experiments for all components, demonstrating that each component possesses unique advantages. Experiment results show that the proposed algorithm outperforms existing schemes in terms of accuracy and latency. The code is available at: https://***/dyxcode/Semi-Asynchronous-Federated-Learning.

关键词： Training Federated learning Heuristic algorithms Predictive models Prediction algorithms Data models Trajectory

来源：评论

学校读者我要写书评

暂无评论

Network Intrusion Detection System Using Deep Learning Method with KDD Cup'99 Dataset 15

Network Intrusion Detection System Using Deep Learning Metho...

引用

15th IEEE International Symposium on Embedded Multicore/Many-Core Systems-on-Chip, MCSoC 2022

作者： Tanimu, Jesse Jeremiah Hamada, Mohamed Robert, Patience Mahendran, Anand Bayero University Kano Dept. of Computer Science Kano Nigeria University of Aizu Software Engineering Lab. Aizu Japan Federal Polytechnic Bali Department of Computer Science Bali Nigeria Higher School of Economics Theoretical Computer Science Lab. Moscow Russia

ISBN: (纸本)9781665464994

This work is a deep sparse autoencoder network intrusion detection system which addresses the issue of interpretability of L2 regularization technique used in other works. The proposed model was trained using a mini-batch gradient descent technique, L1 regularization technique and ReLU activation function to arrive at a better performance. Results based on the KDDCUP'99 dataset show that our approach provides significant performance improvements over other deep sparse autoencoder Network Intrusion Detection Systems. © 2022 IEEE.

关键词： Intrusion detection

来源：评论

学校读者我要写书评

暂无评论

TSEML: A task-specific embedding-based method for few-shot classification of cancer molecular subtypes

arXiv

引用

arXiv 2024年

作者： Su, Ran Shi, Rui Cui, Hui Xuan, Ping Fang, Chengyan Feng, Xikang Jin, Qiangguo School of Computer Software College of Intelligence and Computing Tianjin University Tianjin China School of Software Northwestern Polytechnical University Shaanxi China Department of Computer Science and Information Technology La Trobe University Melbourne Australia Department of Computer Science School of Engineering Shantou University Guangdong China Yangtze River Delta Research Institute of Northwestern Polytechnical University Taicang China

Molecular subtyping of cancer is recognized as a critical and challenging upstream task for personalized therapy. Existing deep learning methods have achieved significant performance in this domain when abundant data samples are available. However, the acquisition of densely labeled samples for cancer molecular subtypes remains a significant challenge for conventional data-intensive deep learning approaches. In this work, we focus on the few-shot molecular subtype prediction problem in heterogeneous and small cancer datasets, aiming to enhance precise diagnosis and personalized treatment. We first construct a new few-shot dataset for cancer molecular subtype classification and auxiliary cancer classification, named TCGA Few-Shot, from existing publicly available datasets. To effectively leverage the relevant knowledge from both tasks, we introduce a task-specific embedding-based meta-learning framework (TSEML). TSEML leverages the synergistic strengths of a model-agnostic meta-learning (MAML) approach and a prototypical network (ProtoNet) to capture diverse and fine-grained features. Comparative experiments conducted on the TCGA Few-Shot dataset demonstrate that our TSEML framework achieves superior performance in addressing the problem of few-shot molecular subtype classification. Copyright © 2024, The Authors. All rights reserved.

关键词： Zero-shot learning

来源：评论

学校读者我要写书评

暂无评论

A Study of Data Augmentation for Imbalance Problem in Network Traffic

A Study of Data Augmentation for Imbalance Problem in Networ...

引用

Blockchain Technology and Information Security (ICBCTIS), International Conference on

作者： Xu Wang Shimin Sun Li Han School of Computer Science and Engineering Tianjin University of Technology Tianjin China Tianjin Key Laboratory of Autonomous Intelligent Technology and System School of Software Tiangong University Tianjin China National Engineering Laboratory for Computer Virus Prevention and Control Technology Tianjin China

The complexity and diversity of smart grid systems increase the likelihood of anomalies in communications between devices in the system, and how these anomalies are detected is critical to the security of the grid system. However, the distribution of traffic types in the grid is often highly unbalanced, with the total amount of normal traffic data often far outweighing the anomalous data. This imbalance compromises the effectiveness of the detection model and poses a significant threat to the overall security of the grid. This paper improves on traditional conditional generation adversarial neural networks and designs a new approach to overcome this obstacle and enhance the security of smart grid systems. In the generator part, the spatial features of the data are captured by convolutional neural networks, and the temporal features of the data are captured by gated recurrent units. By integrating these two components into the generative network, we can generate synthetic data with both temporal and spatial features, thus enhancing the minority class samples. By incorporating the augmented minority data into the training process, the proposed approach enables the detection model to understand the distribution of features of the minority anomaly samples more accurately. This improved learning accuracy translates into improved model performance, resulting in higher detection rates for previously under-represented anomalies. In addition, our approach effectively addresses the inherent bias towards traffic types with higher flows typically observed in asymmetric datasets, thus providing a more equitable and comprehensive security solution for smart grid systems.

关键词：

来源：评论

学校读者我要写书评

暂无评论

SMART: towards pre-trained missing-aware model for patient health status prediction 24

SMART: towards pre-trained missing-aware model for patient h...

引用

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

作者： Zhihao Yu Xu Chu Yujie Jin Yasha Wang Junfeng Zhao School of Computer Science Peking University and National Research and Engineering Center of Software Engineering Peking University School of Computer Science Peking University and Center on Frontiers of Computing Studies Peking University Beijing China and Key Laboratory of High Confidence Software Technologies Ministry of Education Beijing China and Peking University Information Technology Institute (Tianjin Binhai) National Research and Engineering Center of Software Engineering Peking University and Key Laboratory of High Confidence Software Technologies Ministry of Education Beijing China and Peking University Information Technology Institute (Tianjin Binhai) School of Computer Science Peking University and National Research and Engineering Center of Software Engineering Peking University and Key Laboratory of High Confidence Software Technologies Ministry of Education Beijing China and Nanhu Laboratory Jiaxing China

ISBN: (纸本)9798331314385

Electronic health record (EHR) data has emerged as a valuable resource for analyzing patient health status. However, the prevalence of missing data in EHR poses significant challenges to existing methods, leading to spurious correlations and suboptimal predictions. While various imputation techniques have been developed to address this issue, they often obsess difficult-to-interpolate details and may introduce additional noise when making clinical predictions. To tackle this problem, we propose SMART, a Self-Supervised Missing-Aware RepresenTation Learning approach for patient health status prediction, which encodes missing information via missing-aware temporal and variable attentions and learns to impute missing values through a novel self-supervised pre-training approach which reconstructs missing data representations in the latent space rather than in input space as usual. By adopting elaborated attentions and focusing on learning higher-order representations, SMART promotes better generalization and robustness to missing data. We validate the effectiveness of SMART through extensive experiments on six EHR tasks, demonstrating its superiority over state-of-the-art methods. Our code is available at https://***/yzhHoward/SMART.

关键词：

来源：评论

学校读者我要写书评

暂无评论

TSEML: A task-specific embedding-based method for few-shot classification of cancer molecular subtypes

TSEML: A task-specific embedding-based method for few-shot c...

引用

IEEE International Conference on Bioinformatics and Biomedicine (BIBM)

作者： Ran Su Rui Shi Hui Cui Ping Xuan Chengyan Fang Xikang Feng Qiangguo Jin School of Computer Software College of Intelligence and Computing Tianjin University Tianjin China School of Software Northwestern Polytechnical University Shaanxi China Department of Computer Science and Information Technology La Trobe University Melbourne Australia Department of Computer Science School of Engineering Shantou University Guangdong China Yangtze River Delta Research Institute of Northwestern Polytechnical University Taicang China

ISBN: (数字)9798350386226

ISBN: (纸本)9798350386233

Molecular subtyping of cancer is recognized as a critical and challenging upstream task for personalized therapy. Existing deep learning methods have achieved significant performance in this domain when abundant data samples are available. However, the acquisition of densely labeled samples for cancer molecular subtypes remains a significant challenge for conventional data-intensive deep learning approaches. In this work, we focus on the few-shot molecular subtype prediction problem in heterogeneous and small cancer datasets, aiming to enhance precise diagnosis and personalized treatment. We first construct a new few-shot dataset for cancer molecular subtype classification and auxiliary cancer classification, named TCGA Few-Shot, from existing publicly available datasets. To effectively leverage the relevant knowledge from both tasks, we introduce a task-specific embedding-based meta-learning framework (TSEML). TSEML leverages the synergistic strengths of a model-agnostic meta-learning (MAML) approach and a prototypical network (ProtoNet) to capture diverse and fine-grained features. Comparative experiments conducted on the TCGA FewShot dataset demonstrate that our TSEML framework achieves superior performance in addressing the problem of few-shot molecular subtype classification.

关键词： Metalearning Deep learning Knowledge engineering Medical treatment Benchmark testing Bioinformatics Cancer

来源：评论

学校读者我要写书评

暂无评论

A Node Confidence Determination Based Moving Target Defense Strategy for SDN

A Node Confidence Determination Based Moving Target Defense ...

引用

International Conference on Cyber-Enabled Distributed Computing and Knowledge Discovery, CyberC

作者： Shimin Sun Guowei Xu Li Han Tianjin Key Laboratory of Autonomous Intelligent Technology and System School of Software Tiangong Tianjin China School of Computer Science and Engineering Tianjin University of Technology National Engineering Laboratory for Computer Virus Prevention and Control Technology Tianjin China

With the rapid development of Internet technology, network assets have become an indispensable part of modern society. Especially in the face of high-value network assets, the attacker is faced with paying a greater cost to study the network structure. Due to the unknown attacker and attack means, the attacker has an advantage in the attack and defense situation. The moving target defense technique can flexibly change the static characteristics of the network attack surface and increase the difficulty for the attacker to detect the network, thus achieving the purpose of enhancing network security. Based on this technique, this paper proposes a defense strategy based on node confidence. It puts forward stricter constraints on the differential selection of address/port hopping paths, dynamically selects the hopping objects according to the node’s confidence, and implements the defense strategy of differential complexity. This further increases the network’s uncertainty and the difficulty for the attacker to detect the network structure. The experimental results show that in the path differentiation comparison, the maximum improvement is about 18% compared to the Dynamic Random Route Hopping strategy. In the CPU load experiment, the CPU load rate is reduced by about 4.625% on average compared to the Double Address Hopping strategy. Moreover, for larger hopping period, the proposed strategy also performs well in terms of CPU load rate. The proposed strategy also exhibits good performance in terms of CPU load ratio for large hopping cycles.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Towards Detecting Comment-Code Inconsistency in Smart Contract via Pre-Training Techniques

Towards Detecting Comment-Code Inconsistency in Smart Contra...

引用

2024 IEEE International Conference on Security, Privacy, Anonymity in Computation and Communication and Storage (SpaCCS)

作者： Xiaohui Liu Sicheng Hao Lei Yang Xiaohong Shi School of Computer Science and Engineering Sun Yat-sen University Guangzhou China School of Software Engineering Sun Yat-sen University Zhuhai China School of Information Engineering Guangzhou Panyu Polytechnic Guangzhou China

ISBN: (数字)9798331521073

ISBN: (纸本)9798331521080

Smart contracts are programs deployed on the blockchain. The security issues of smart contracts have become the focus in recent years. Comments have been widely used to improve the readability of programs. Nevertheless, comments and code are not always consistent in smart contracts because of the modification of developers based on token standards and code reuse. These inconsistencies can mislead developers and even pose security risks. We call comment-code inconsistencies as CCIs. In particular, permission-related CCIs can easily expose vulnerabilities in smart contracts, causing funds to be hacked. In this work, we employ pre-training techniques to detect CCIs on permission in smart contracts. Specifically, we collect 3511 comments from Ethereum and Defi projects. Firstly, we utilize BERT to classify permission comments. Then, we extract related function implementations of each permission comment using the compilation technique and adopt CodeT5 to embed comment-code pairs. Finally, we make use of BiLSTM to predict inconsistencies. The evaluation results show that our proposed approach performs better than the existing tools in both precision and recall. Moreover, considering the two stages, it achieves an accuracy of 95.2% and 91.5% in permission comment recognition and permission CCI detection, respectively.

关键词：

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：