检索结果-内蒙古大学图书馆

arXiv 2023年

作者： Montaruli, Biagio Floris, Giuseppe Scano, Christian Demetrio, Luca Valenza, Andrea Compagna, Luca Ariu, Davide Piras, Luca Balzarotti, Davide Biggio, Battista Dept. of Digital Security EURECOM Biot06410 France Dept. of Electrical and Electronic Engineering University of Cagliari Cagliari09124 Italy Department of Computer Control and Management Engineering Sapienza University Rome Italy University of Genova Genova16146 Italy Prima Assicurazioni Milano20131 Italy Endor Labs United States Pluribus One Cagliari09128 Italy

Many Web Application Firewalls (WAFs) leverage the OWASP Core Rule Set (CRS) to block incoming malicious requests. The CRS consists of different sets of rules designed by domain experts to detect well-known web attack patterns. Both the set of rules to be used and the weights used to combine them are manually defined, yielding four different default configurations of the CRS. In this work, we focus on the detection of SQL injection (SQLi) attacks, and show that the manual configurations of the CRS typically yield a suboptimal trade-off between detection and false alarm rates. Furthermore, we show that these configurations are not robust to adversarial SQLi attacks, i.e., carefully-crafted attacks that iteratively refine the malicious SQLi payload by querying the target WAF to bypass detection. To overcome these limitations, we propose (i) using machine learning to automate the selection of the set of rules to be combined along with their weights, i.e., customizing the CRS configuration based on the monitored web services;and (ii) leveraging adversarial training to significantly improve its robustness to adversarial SQLi manipulations. Our experiments, conducted using the well-known open-source ModSecurity WAF equipped with the CRS rules, show that our approach, named ModSec-AdvLearn, can (i) increase the detection rate up to 30%, while retaining negligible false alarm rates and discarding up to 50% of the CRS rules;and (ii) improve robustness against adversarial SQLi attacks up to 85%, marking a significant stride toward designing more effective and robust WAFs. We release our open-source code at https://***/pralab/modsec-advlearn. © 2023, CC BY-SA.

关键词： Adversarial machine learning

来源：评论

学校读者我要写书评

暂无评论

Feedforward control in the Presence of Input Nonlinearities: A Learning-based Approach

arXiv

引用

arXiv 2022年

作者： van Hulst, Jilles Poot, Maurice Kostić, Dragan Yan, Kai Wa Portegies, Jim Oomen, Tom Control Systems Technology Section Dept. of Mechanical Engineering Eindhoven University of Technology Netherlands Center of Competency ASM Pacific Technology Netherlands CASA Dept. of Mathematics and Computer Science Eindhoven University of Technology Netherlands Delft Center for Systems and Control Delft University of Technology Netherlands

Advanced feedforward control methods enable mechatronic systems to perform varying motion tasks with extreme accuracy and throughput. The aim of this paper is to develop a data-driven feedforward controller that addresses input nonlinearities, which are common in typical applications such as semiconductor back-end equipment. The developed method consists of parametric inverse-model feedforward that is optimized for tracking error reduction by exploiting ideas from iterative learning control. Results on a simulated set-up indicate improved performance over existing identification methods for systems with nonlinearities at the input. © 2022, CC BY-NC-ND.

关键词： Feedforward control

来源：评论

学校读者我要写书评

暂无评论

A baseline accuracy classification method to over-come the over-fitting problem for class- imbalanced defect-prone datasets model 14

A baseline accuracy classification method to over-come the o...

引用

14th Annual IEEE International Systems Conference, SYSCON 2020

作者： Shaikh, Salahuddin Changan, Liu Malik, Maaz Rasheed Dept. School of Control and Computer Engineering North China Electric Power University Beijing China Dept. Information and Communication Engineering Guilin University of Electronics Technology Guilin China

ISBN: (纸本)9781728153650

Software Engineers are battling always with defects from the earliest starting point of the software history. The defect has significant commitment and assumes an important job in the hour of freeing software. Software defect-prone might be recognized before the software announcement or after the announcement. Although, the desire for designers is to give defect free software to their customers. Hence, recognizing the software defect-prone before announcement makes a decent preferred position for the designers. Here, the difficult assignment is to discover the defect-prone region or defective module of software. On the off chance that defective modules are distinguished before the arrival of software, at that point it could be effectively counteracted and fixed by the designer. In most recent two decades scientists have concentrated on software deformity prophecy issue by applying a few statistical and machine learning systems. The software defect-prone information experiences the class-imbalance issue due to the slanted conveyance of defective and non-defective software modules. Generally, machine learning calculations consider equivalent circulation of information tests in each class and expect the misclassification cost of each class is similarly significant. We have used SMOTE with ONE-R by using its MinBucketsize number n=1, 2, 3, 4, 5 & 6 for over-fit the data and resultant we got that minbucketsize n= 1 & 2 are highly good for base accuracy and efficiency for defect-prone datasets model. We have used datasets model in three different ways and we observed that use training datasets is very suitable for our experiments as compare to other two ways. We also observed that al evaluation measures results are highly good at use of training datasets model. © 2020 IEEE

关键词： Classification (of information)

来源：评论

学校读者我要写书评

暂无评论

The role of attribute ranker using classification for software defect-prone data-sets model: An empirical comparative study 14

The role of attribute ranker using classification for softwa...

引用

14th Annual IEEE International Systems Conference, SYSCON 2020

作者： Malik, Maaz Rasheed Yining, Liu Shaikh, Salahuddin Dept. Information and Communication Engineering Guilin University of Electronics Technology Guilin China Dept. School of Control and Computer Engineering North China Electric Power University Beijing China

ISBN: (纸本)9781728153650

Feature selection, is an issue firmly identified with size decrease of data-sets model. The target of feature selection is to recognize features in the data-set as significant, and dispose of some other feature as unimportant and repetitive data. Since feature selection diminishes the dimensionality of the data-sets model, it holds out the probability of increasingly successful and quick activity of data mining algorithm which can be worked quicker and all the more adequately by utilizing feature selection. In this research paper we will investigate feature extraction Principal Component Analysis (PCA) with attribute ranker search technique. In practice, Principal Component Analysis with attribute ranker search strategy isn't just used to improve extra storage space or the computational accuracy and efficiency of the classification algorithm, however can likewise enhance the prescient presentation by diminishing the scourge of dimensionality - particularly on the off chance that we are working with software defect-prone or non-defected data-sets models. We have used 10 datasets models, these datasets models basically REPOSITORY model of NASA which contain binary class defected and non-defected datasets models. We have also used 6 classifiers for comparatively analysis between objective model and real datasets model. We illustrated the comparatively analysis between PCA Ranker Search method with No-PCA. We have also compared classifiers efficiency with each other. The most efficient and useful classifier are the Bagging and Multilayer Perceptron at all attribute ranking search method. But the comparatively analysis between the classifiers that Naïve bayes and MultiLayer Perceptron have well increased the correctly classified instances percentage in overall software fault forecast. One more comparison between the PCA Ranker Search Method and No-PCA that is attribute ranker search method is really good for increasing accuracy and efficiency for software fault forecast

关键词： Defects

来源：评论

学校读者我要写书评

暂无评论

GFT: Gradient Focal Transformer

arXiv

引用

arXiv 2025年

作者： Kriuk, Boris Gill, Simranjit Kaur Aslam, Shoaib Fakhrutdinov, Amir Hong Kong University of Science and Technology Dept. of Computer Science & Engineering Hong Kong University of Westminster Dept. of Computer Science & Engineering United Kingdom University of Engineering & Technology Lahore Dept. of Mechanical Mechatronics & Control Eng Pakistan Shanghai Jiao Tong University Dept. of Mechanical Engineering China

Fine-Grained Image Classification (FGIC) remains a complex task in computer vision, as it requires models to distinguish between categories with subtle localized visual differences. Well-studied CNN-based models, while strong in local feature extraction, often fail to capture the global context required for fine-grained recognition, while more recent ViT-backboned models address FGIC with attention-driven mechanisms but lack the ability to adaptively focus on truly discriminative regions. TransFG and other ViT-based extensions introduced part-aware token selection to enhance attention localization, yet they still struggle with computational efficiency, attention region selection flexibility, and detail-focus narrative in complex environments. This paper introduces GFT (Gradient Focal Transformer), a new ViT-derived framework created for FGIC tasks. GFT integrates the Gradient Attention Learning Alignment (GALA) mechanism to dynamically prioritize class-discriminative features by analyzing attention gradient flow. Coupled with a Progressive Patch Selection (PPS) strategy, the model progressively filters out less informative regions, reducing computational overhead while enhancing sensitivity to fine details. GFT achieves SOTA accuracy on FGVC Aircraft, Food-101, and COCO datasets with 93M parameters, outperforming ViT-based advanced FGIC models in efficiency. By bridging global context and localized detail extraction, GFT sets a new benchmark in fine-grained recognition, offering interpretable solutions for real-world deployment scenarios. Copyright © 2025, The Authors. All rights reserved.

关键词： Deep learning

来源：评论

学校读者我要写书评

暂无评论

Polytopic input constraints in learning-based optimal control using neural networks

arXiv

引用

arXiv 2021年

作者： Markolf, Lukas Stursberg, Olaf Control and System Theory Group Dept. of Electrical Engineering and Computer Science University of Kassel Germany

This work considers artificial feed-forward neural networks as parametric approximators in optimal control of discrete-time systems. Two different approaches are introduced to take polytopic input constraints into account. The first approach determines (sub-)optimal inputs by the application of gradient methods. Closed-form expressions for the gradient of general neural networks with respect to their inputs are derived. The approach allows to consider state-dependent input constraints, as well as to ensure the satisfaction of state constraints by exploiting recursive reachable set computations. The second approach makes use of neural networks with softmax output units to map states into parameters, which determine (sub-)optimal inputs by a convex combination of the vertices of the input constraint set. The application of both approaches in model predictive control is discussed, and results obtained for a numerical example are used for illustration. Copyright © 2021, The Authors. All rights reserved.

关键词： Discrete time control systems

来源：评论

学校读者我要写书评

暂无评论

Pinching Antennas: Principles, Applications and Challenges

arXiv

引用

arXiv 2025年

作者： Yang, Zheng Wang, Ning Sun, Yanshi Ding, Zhiguo Schober, Robert Karagiannidis, George K. Wong, Vincent W.S. Dobre, Octavia A. Fujian Provincial Engineering Technology Research Center of Photoelectric Sensing Application Fujian Normal University Fuzhou350117 China School of Information and Control Engineering China University of Mining and Technology Xuzhou221116 China School of Computer Science and Information Engineering Hefei University of Technology Hefei230009 China Khalifa University Abu Dhabi United Arab Emirates University of Manchester ManchesterM1 9BB United Kingdom Germany Group Electrical & Computer Engineering Dept. Aristotle University of Thessaloniki Thessaloniki54 124 Greece Department of Electrical and Computer Engineering The University of British Columbia Vancouver Canada Department of Electrical and Computer Engineering Memorial University St. John’sNL Canada

Flexible-antenna systems, such as fluid antennas and movable antennas, have been recognized as key enabling technologies for sixth-generation (6G) wireless networks, as they can intelligently reconfigure the effective channel gains of the users and hence significantly improve their data transmission capabilities. However, existing flexible-antenna systems have been designed to combat small-scale fading in non-line-of-sight (NLoS) conditions. As a result, they lack the ability to establish line-of-sight links, which are typically 100 times stronger than NLoS links. In addition, existing flexible-antenna systems have limited flexibility, where adding/removing an antenna is not straightforward. This article introduces an innovative flexible-antenna system called pinching antennas, which are realized by applying small dielectric particles to waveguides. We first describe the basics of pinching-antenna systems and their ability to provide strong LoS links by deploying pinching antennas close to the users as well as their capability to scale up/down the antenna system. We then focus on communication scenarios with different numbers of waveguides and pinching antennas, where innovative approaches to implement multiple-input multiple-output and non-orthogonal multiple access are discussed. In addition, promising 6G-related applications of pinching antennas, including integrated sensing and communication and next-generation multiple access, are presented. Finally, important directions for future research, such as waveguide deployment and channel estimation, are highlighted. Copyright © 2025, The Authors. All rights reserved.

关键词： Channel estimation

来源：评论

学校读者我要写书评

暂无评论

Graph neural networks to model and optimize the operation of Water Distribution Networks: A review

引用

Journal of Industrial Information Integration 2025年

作者： Giacomo Vittori Yelizaveta Falkouskaya Daniel M. Jimenez-Gutierrez Tiziana Cattai Ioannis Chatzigiannakis Dept. of Information Engineering Electronics and Telecommunications Sapienza University of Rome Rome Italy Dept. of Computer Control and Management Engineering Sapienza University of Rome Rome Italy

Water Distribution Networks (WDNs) have become increasingly complex and interconnected, and the need for advanced modeling and optimization techniques has become fundamental to ensure an efficient and reliable clean water supply. Representing WDNs as graphs naturally models the underlying interacting physical structure and enables the usage of Graph Neural Networks (GNN) that combine the physical structure with abstract notions to capture local and global relationships. GNNs offer significant advantages in contrast to generic Deep Learning (DL) techniques and stand out as a promising solution to model intricate dependencies and enable the investigation of key challenges such as leak detection, water quality monitoring, and demand forecasting. This review presents the physics and hydraulics involved in WDN and the prevalent graph-based models used in the literature. The theoretical foundations of GNNs are shown, highlighting their capabilities in capturing complex spatial relationships and dependencies inherent in the network topology. The most promising GNN-based solutions that can address some of the most critical challenges of WDNs are discussed in detail. We outline the open challenges and potential directions for future developments in this field. By combining multidisciplinary and real-world aspects, this critical review highlights the role of GNNs in modeling and optimizing WDNs.

关键词： Water Distribution Networks Graph neural network Graph theory Network planning Network operation Leak detection Demand forecasting Optimal sensor placement Water quality monitoring

来源：评论

学校读者我要写书评

暂无评论

The Goldilocks Framework: Towards Selecting the Optimal Approach to Conducting AI Projects

The Goldilocks Framework: Towards Selecting the Optimal Appr...

引用

AI engineering - Software engineering for AI (WAIN), IEEE/ACM Workshop on

作者： Rimman Dzhusupova Jan Bosch Helena Holmströ m Olsson Dept. of Electrical &#x0026 Instrumentation Control &#x0026 Safety Systems McDermott The Hague The Netherlands Department of Computer Science and Engineering Chalmers University of Technology Gothenburg Sweden Department of Computer Science and Media Technology Malm&#x00F6 University Malm&#x00F6 Sweden

Artificial intelligence is increasingly becoming important to businesses since many companies have realized the benefits of applying Machine Learning (ML) and Deep Learning (DL) into their operations. Nevertheless, ML/DL technologies’ industrial development and deployment examples are still rare and generally confined within a small cluster of large international companies who are struggling to apply ML more broadly and deploy their use cases at a large scale. Meanwhile, current AI market has started offering various solutions and services. Thus, organizations must understand how to acquire AI technology based on their business strategy and available resources. This paper discusses the industrial experience of developing and deploying ML/DL use cases to support organizations in their transformation towards AI. We identify how various factors, like cost, schedule, and intellectual property, can be affected by the choice of approach towards ML/DL project development and deployment within large international engineering corporations. As a research result, we present a framework that covers the trade-offs between those various factors and can support engineering companies to choose the best approach based on their long-term business strategies and, therefore, would help to accomplish their ML/DL project deployment successfully.

关键词：

来源：评论

学校读者我要写书评

暂无评论

FDP-GAMLP: New Financial Distress Prediction Technique based on Genetic Algorithm and Multi-Layer Perceptron 19

FDP-GAMLP: New Financial Distress Prediction Technique based...

引用

19th IEEE International Multi-Conference on Systems, Signals and Devices, SSD 2022

作者： Khedr, Ahmed M. Bannany, Magdi El Saif, Abdul-Wahid A. University of Sharjah Computer Science Dept. Sharjah27272 United Arab Emirates College of Business Administration Umm Al Quwain University Umm Al Quwain356 United Arab Emirates Ain Shams University Faculty of Business Cairo11566 Egypt King Fahd University of Petroleum & Minerals Control and Instrumentation Engineering Department Dhahran Saudi Arabia

ISBN: (数字)9781665471084

ISBN: (纸本)9781665471084

Developing a reliable financial distress prediction model has been a long-standing research area. Recently, machine learning algorithms have been increasingly popular in the area. In this paper, a hybrid approach of Genetic Algorithm (GA) and Multi-Layer Perceptron (MLP) for Financial Distress Prediction (FDP) (FDP-GAMLP) is proposed. FDP-GAMLP emphasizes on GA-based tuning of the four major hyper-parameters, namely network width, network dept., network optimizer, and dense layer activation function, which can influence on whether the algorithm explodes or converges. An improved GA is utilized to optimize the MLP model's hyper-parameters for better prediction. The prediction performance is evaluated using real data set with samples of companies from countries in Middle East region. The resampling technique using k-fold evaluation metrics is adopted to get unbiased and most accurate results. The simulation results show that FDP-GAMLP outperforms the classical machine learning models in terms of predictive accuracy. © 2022 IEEE.

关键词： Genetic algorithms

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：