检索结果-内蒙古大学图书馆

PPO Based Reinforcement Learning Horizontal Pods Auto-Scaling-A Case Study in Simulation environment

学校读者我要写书评

暂无评论

PPO Based Reinforcement Learning Horizontal Pods Auto-Scalin...

Advances in Electrical Engineering and Computer Applications( AEECA), 2020 IEEE International Conference on

作者： Fang Peng Xin Ji Le Zhang Ruibo Chen Xingchuang Liao Kui Zhang Big Data Center State Grid Corporation of China Beijing China State Key Laboratory of Software Development Environment Beihang University Beijing China

Microservice has become a state-of-art architecture in the software engineering domain. In the big data scenario, the data can be encapsulated by microservice, providing independent service or orchestrating the business procedure. Much research, such as threshold-based, rule-based, and even deep learning-based resource scaling, has been proposed to tackle this problem. However, the model trained on data like log or monitor index limited the intended dynamic resource adjusting, which caused high latency during adjusting. Moreover, the computation complexity and the data latency can still be a hindrance to the problem-solving. This paper proposed a reinforcement learning-based horizontal pods auto-scaling algorithm to study and respond in low latency. Furthermore, a case study to validate the effectiveness was carried out in a simulation environment. The results show that the proposed algorithm can converge after 0.3 million steps, and three groups of experiments were carried out, which shows that the trained reinforcement learning model can adjust the replica number in real-time, which indicates that the proposed algorithm can be applied in a scenario with low resource consumption.

关键词：

Attack End-to-End Autonomous Driving through Module-Wise Noise

学校读者我要写书评

暂无评论

arXiv 2024年

作者： Wang, Lu Zhang, Tianyuan Han, Yikai Fang, Muyang Jin, Ting Kang, Jiaqi School of Computer Science and Engineering Beihang University Beijing China Shen Yuan Honors College Beihang University Beijing China State Key Lab of Software Development Environment Beihang University Beijing China School of Software Beihang University Beijing China

With recent breakthroughs in deep neural networks, numerous tasks within autonomous driving have exhibited remarkable performance. However, deep learning models are susceptible to adversarial attacks, presenting significant security risks to autonomous driving systems. Presently, end-to-end architectures have emerged as the predominant solution for autonomous driving, owing to their collaborative nature across different tasks. Yet, the implications of adversarial attacks on such models remain relatively unexplored. In this paper, we conduct comprehensive adversarial security research on the modular end-to-end autonomous driving model for the first time. We thoroughly consider the potential vulnerabilities in the model inference process and design a universal attack scheme through module-wise noise injection. We conduct large-scale experiments on the full-stack autonomous driving model and demonstrate that our attack method outperforms previous attack methods. We trust that our research will offer fresh insights into ensuring the safety and reliability of autonomous driving systems. © 2024, CC BY-NC-ND.

关键词： Deep neural networks

Noise-injected Consistency Training and Entropy-constrained Pseudo labeling for Semi-supervised Extractive Summarization 29

学校读者我要写书评

暂无评论

Noise-injected Consistency Training and Entropy-constrained ...

29th International Conference on Computational Linguistics, COLING 2022

作者： Wang, Yiming Mao, Qianren Liu, Junnan Jiang, Weifeng Zhu, Hongdong Li, Jianxin Beijing Advanced Innovation Center for Big Data and Brain Computing Beihang University China The State Key Laboratory of Software Development Environment Beihang University China Institute of Artificial Intelligence Beihang University China

labeling large amounts of extractive summarization data is often prohibitive expensive due to time, financial, and expertise constraints, which poses great challenges to incorporating summarization system in practical applications. This limitation can be overcome by semi-supervised approaches: consistency-training and pseudo-labeling to make full use of unlabeled data. Researches on the two, however, are conducted independently, and very few works try to connect them. In this paper, we first use the noise-injected consistency training paradigm to regularize model predictions. Subsequently, we propose a novel entropy-constrained pseudo labeling strategy to obtain high-confidence labels from unlabeled predictions, which can obtain high-confidence labels from unlabeled predictions by comparing the entropy of supervised and unsupervised predictions. By combining consistency training and pseudo-labeling, this framework enforce a low-density separation between classes, which decently improves the performance of supervised learning over an insufficient labeled extractive summarization dataset. © 2022 Proceedings - International Conference on Computational Linguistics, COLING. All rights reserved.

关键词： Entropy

Deep Fused Flow and Topology Features for Botnet Detection Basing on Pretrained Gcn

学校读者我要写书评

暂无评论

SSRN

SSRN 2023年

作者： Meng, Xiaoyuan Lang, Bo Liu, Yanxi Yan, Yuhao State Key Laboratory of Software Development Environment Beihang University XueYuan Road Beijing100191 China Zhongguancun Laboratory Beijing China

Botnets become significant cyber security threats nowadays. The characteristics of botnets are mainly reflected in the network behaviors of the bots and their intercommunication relationships. Existing botnet detection methods usually only use one kind of features, i.e., flow features or topology features, which overlooks the other type of features and affects the model performance. In this paper, we propose a botnet detection model which uses graph convolutional network (GCN) to deeply fuse flow features and topology features for the first time. We construct communication graphs from network traffics and represent nodes attributes with flow features. The extreme sample imbalance problem in the existing public traffic flow datasets makes it impractical to train a GCN model on these datasets. Therefore, we use a public balanced communication graph dataset to pretrain the GCN model, which equips the model with the ability to identify botnet topology properties. When fed with a communication graph, the feature output from the last hidden layer of the GCN model fuses the flow and the topology information. Furthermore, by adjusting the number of layers in the GCN network, our model can achieve effective Command-and-Control (C2) and Peer-to-Peer (P2P) botnets detections. Experimental results on public datasets demonstrate that our approach outperforms the current state-of-the-art botnet detection models. © 2023, The Authors. All rights reserved.

关键词： Botnet

PriSTI: A Conditional Diffusion Framework for Spatiotemporal Imputation

学校读者我要写书评

暂无评论

PriSTI: A Conditional Diffusion Framework for Spatiotemporal...

International Conference on Data Engineering

作者： Mingzhe Liu Han Huang Hao Feng Leilei Sun Bowen Du Yanjie Fu State Key Laboratory of Software Development Environment Beihang University Beijing China Department of Computer Science University of Central Florida USA

Spatiotemporal data mining plays an important role in air quality monitoring, crowd flow modeling, and climate forecasting. However, the originally collected spatiotemporal data in real-world scenarios is usually incomplete due to sensor failures or transmission loss. Spatiotemporal imputation aims to fill the missing values according to the observed values and the underlying spatiotemporal dependence of them. The previous dominant models impute missing values autoregressively and suffer from the problem of error accumulation. As emerging powerful generative models, the diffusion probabilistic models can be adopted to impute missing values conditioned by observations and avoid inferring missing values from inaccurate historical imputation. However, the construction and utilization of conditional information are inevitable challenges when applying diffusion models to spatiotemporal imputation. To address above issues, we propose a conditional diffusion framework for spatiotemporal imputation with enhanced prior modeling, named PriSTI. Our proposed framework provides a conditional feature extraction module first to extract the coarse yet effective spatiotemporal dependencies from conditional information as the global context prior. Then, a noise estimation module transforms random noise to realistic values, with the spatiotemporal attention weights calculated by the conditional feature, as well as the consideration of geographic relationships. PriSTI outperforms existing imputation methods in various missing patterns of different real-world spatiotemporal data, and effectively handles scenarios such as high missing rates and sensor failure. The implementation code is available at https://***/LMZZML/PriSTI.

关键词：

label-Enhanced Graph Neural Network for Semi-supervised Node Classification

学校读者我要写书评

暂无评论

arXiv 2022年

作者： Yu, Le Sun, Leilei Du, Bowen Zhu, Tongyu Lv, Weifeng The State Key Laboratory of Software Development Environment Beihang University Beijing100191 China

Graph Neural Networks (GNNs) have been widely applied in the semi-supervised node classification task, where a key point lies in how to sufficiently leverage the limited but valuable label information. Most of the classical GNNs solely use the known labels for computing the classification loss at the output. In recent years, several methods have been designed to additionally utilize the labels at the input. One part of the methods augment the node features via concatenating or adding them with the one-hot encodings of labels, while other methods optimize the graph structure by assuming neighboring nodes tend to have the same label. To bring into full play the rich information of labels, in this paper, we present a label-enhanced learning framework for GNNs, which first models each label as a virtual center for intra-class nodes and then jointly learns the representations of both nodes and labels. Our approach could not only smooth the representations of nodes belonging to the same class, but also explicitly encode the label semantics into the learning process of GNNs. Moreover, a training node selection technique is provided to eliminate the potential label leakage issue and guarantee the model generalization ability. Finally, an adaptive self-training strategy is proposed to iteratively enlarge the training set with more reliable pseudo labels and distinguish the importance of each pseudo-labeled node during the model training process. Experimental results on both real-world and synthetic datasets demonstrate our approach can not only consistently outperform the state-of-the-arts, but also effectively smooth the representations of intra-class nodes. Copyright © 2022, The Authors. All rights reserved.

关键词： Graph neural networks

Variant quantifiers in L_(3)-valued first-order logic

学校读者我要写书评

暂无评论

Frontiers of Computer Science 2021年第5期15卷 151-159页

作者： Wei LI Yuefei SUI State Key Laboratory of Software Development Environment Beihang UniversityBeijing 100083China Key Laboratory of Intelligent Information Processing Institute of Computing TechnologyChinese Academy of SciencesBeijing 100190China School of Computer Science and Technology University of Chinese Academy of SciencesBeijing 100049China

Traditional first-order logic has four definitions for quantifiers,which are defined by universal and existential *** L_(3)-valued(three-valued)first-order logic,there are eight kinds of definitions for quantifiers;and corresponding Gentzen deduction systems will be given and their soundness and completeness theorems will be proved.

关键词： L_(3)-valued(three-valued)first-order logic sequent quantifier soundness theorem completeness theorem

Defensive Patches for Robust Recognition in the Physical World

学校读者我要写书评

暂无评论

arXiv 2022年

作者： Wang, Jiakai Yin, Zixin Hu, Pengfei Liu, Aishan Tao, Renshuai Qin, Haotong Liu, Xianglong Tao, Dacheng State Key Lab of Software Development Environment Beihang University Beijing China JD Explore Academy Beijing China

To operate in real-world high-stakes environments, deep learning systems have to endure noises that have been continuously thwarting their robustness. Data-end defense, which improves robustness by operations on input data instead of modifying models, has attracted intensive attention due to its feasibility in practice. However, previous data-end defenses show low generalization against diverse noises and weak transferability across multiple models. Motivated by the fact that robust recognition depends on both local and global features, we propose a defensive patch generation framework to address these problems by helping models better exploit these features. For the generalization against diverse noises, we inject class-specific identifiable patterns into a confined local patch prior, so that defensive patches could preserve more recognizable features towards specific classes, leading models for better recognition under noises. For the transferability across multiple models, we guide the defensive patches to capture more global feature correlations within a class, so that they could activate model-shared global perceptions and transfer better among models. Our defensive patches show great potentials to improve application robustness in practice by simply sticking them around target objects. Extensive experiments show that we outperform others by large margins (improve 20+% accuracy for both adversarial and corruption robustness on average in the digital and physical world). Copyright © 2022, The Authors. All rights reserved.

关键词： Deep learning

Multilingual Entity and Relation Extraction from Unified to Language-specific Training

学校读者我要写书评

暂无评论

arXiv 2023年

作者： Wang, Zixiang Yang, Jian Li, Tongliang Liu, Jiaheng Mo, Ying Bai, Jiaqi He, Longtao Li, Zhoujun State Key Lab of Software Development Environment Beihang University Beijing China National Computer Network Emergency Response Technical Team Coordination Center of China Beijing China

Entity and relation extraction is a key task in information extraction, where the output can be used for downstream NLP tasks. Existing approaches for entity and relation extraction tasks mainly focus on the English corpora and ignore other languages. Thus, it is critical to improving performance in a multilingual setting. Meanwhile, multilingual training is usually used to boost cross-lingual performance by transferring knowledge from languages (e.g., high-resource) to other (e.g., low-resource) languages. However, language interference usually exists in multilingual tasks as the model parameters are shared among all languages. In this paper, we propose a two-stage multilingual training method and a joint model called Multilingual Entity and Relation Extraction framework (mERE) to mitigate language interference across languages. Specifically, we randomly concatenate sentences in different languages to train a Language-universal Aggregator (LA), which narrows the distance of embedding representations by obtaining the unified language representation. Then, we separate parameters to mitigate interference via tuning a Language-specific Switcher (LS), which includes several independent sub-modules to refine the language-specific feature representation. After that, to enhance the relational triple extraction, the sentence representations concatenated with the relation feature are used to recognize the entities. Extensive experimental results show that our method outperforms both the monolingual and multilingual baseline methods. Besides, we also perform detailed analysis to show that mERE is lightweight but effective on relational triple extraction and mERE is easy to transfer to other backbone models of multi-field tasks, which further demonstrates the effectiveness of our method. Copyright © 2023, The Authors. All rights reserved.

关键词： Information retrieval