检索结果-内蒙古大学图书馆

Dual-filtering (DF) schemes for learning systems to prevent adversarial attacks

COMPLEX & INTELLIGENT SYSTEMS 2023年第4期9卷 3717-3738页

作者： Dasgupta, Dipankar Gupta, Kishor Datta Univ Memphis Memphis TN 38152 USA

Defenses against adversarial attacks are essential to ensure the reliability of machine-learning models as their applications are expanding in different domains. Existing ML defense techniques have several limitations in practical use. We proposed a trustworthy framework that employs an adaptive strategy to inspect both inputs and decisions. In particular, data streams are examined by a series of diverse filters before sending to the learning system and then crossed checked its output through anomaly (outlier) detectors before making the final decision. Experimental results (using benchmark data-sets) demonstrated that our dual-filtering strategy could mitigate adaptive or advanced adversarial manipulations for wide-range of ML attacks with higher accuracy. Moreover, the output decision boundary inspection with a classification technique automatically affirms the reliability and increases the trustworthiness of any ML-based decision support system. Unlike other defense techniques, our dual-filtering strategy does not require adversarial sample generation and updating the decision boundary for detection, makes the ML defense robust to adaptive attacks.

关键词： Adversarial machine learning Computer security Bio-inspired algorithm negative selection algorithm

来源：评论

学校读者我要写书评

暂无评论

A negative selection algorithm based on hierarchical clustering of self set

引用

Science China(Information Sciences) 2013年第8期56卷 203-215页

作者： CHEN Wen LI Tao LIU XiaoJie ZHANG Bing College of Computer Science Sichuan University National Computer Network Emergency Response Technical Team Coordination Center of China

negative selection algorithm(NSA) is an important method of generating artificial immune ***,the traditional NSAs aim at eliminating the self-recognized invalid detectors,by matching candidate detectors with the whole self *** matching process results in extremely low generation efficiency and significantly limits the application of *** this paper,an improved NSA called CB-RNSA,which is based on the hierarchical clustering of self set,is *** CB-RNSA,the self data is first preprocessed by hierarchical clustering,and then replaced by the self cluster centers to match with candidate detectors in order to reduce the distance calculation *** the detector generation process,the candidate detectors are restricted to the lower coverage space to reduce the detector *** the paper,probabilistic analysis is performed on non-self coverage of ***,termination condition of the detector generation procedure in CB-RNSA is *** is more reasonable than that of traditional NSAs,which are based on predefined detector *** theoretical analysis shows the time complexity of CB-RNSA is irrelevant to the self set ***,the difficult problem,in which the detector training cost is exponentially related to the size of self set in traditional NSAs,is resolved,and the efficiency of the detector generation under a big self set is also *** experimental results show that:under the same data set and expected coverage,the detection rate of CB-RNSA is higher than that of the classic RNSA and V-detector algorithms by 12.3% and 7.4% ***,the false alarm rate is lower by 8.5% and 4.9% respectively,and the time cost of CB-RNSA is lower by 67.6% and 75.7% respectively.

关键词： artificial immune system negative selection algorithm detector cluster

来源：评论

学校读者我要写书评

暂无评论

Anomaly Detection Using a Novel negative selection algorithm

引用

JOURNAL OF COMPUTATIONAL AND THEORETICAL NANOSCIENCE 2013年第12期10卷 2831-2835页

作者： Zeng, Jinquan Qin, Zhiguang Tang, Weiwen Univ Elect Sci & Technol China Sch Comp Sci & Engn Chengdu 610054 Peoples R China Sichuan Commun Res Planning & Designing Co Ltd Chengdu 610041 Peoples R China

negative selection algorithm (NSA) is one of the major algorithms developed within artificial immune system (AIS) and can be used for network security, fault detection, especially, anomaly detection. NSA generates the detectors based on the self space. Due to the drawbacks of the current representation of the self space in NSAs, the generated detectors cannot enough cover the non-self space and at the same time, cover some of the self space. In this paper, an extension of real-valued negative selection algorithm with the variable-sized self radius and bi-directional matching rule are introduced. Using the variable-sized self radius and bi-directional matching rule, we can construct an appropriate profile of the system, and then based on the constructed profile of the system we produce the more quality detectors to cover the non-self space;at the same time, using the variable-sized self radius and bi-directional matching rule, we can decrease the number of detectors and cover enough the non-self space. In our experiments, this approach is tested using the well-known real world datasets;preliminary results show that the new approach improves the overall performance of detectors and without increase in complexity.

关键词： Artificial Immune Systems negative selection algorithm Anomaly Detection

来源：评论

学校读者我要写书评

暂无评论

NKA: a pathogen dose-based natural killer cell algorithm and its application to classification

引用

JOURNAL OF SUPERCOMPUTING 2022年第5期78卷 7016-7037页

作者： Wang, Dongmei Liang, Yiwen Yang, Xinmin Wuhan Univ Sch Comp Sci Wuhan 430072 Peoples R China Wuhan Univ Sch Cyber Sci & Engn Wuhan 430072 Peoples R China

Inspired by the self/non-self discrimination theory, Forrest et al. proposed a distance-based negative selection algorithm (NSA). However, in the detector generation phase, NSA creates specific antibodies for each antigen based on the mechanism of specific antigen-antibody binding reactions, which lead to an excessive and redundant number of detectors. Moreover, during the detection phase, NSA needs to calculate the matching degree of the antigen with all the detectors, which has a high computational cost and low detection efficiency. This letter presents a novel natural killer cell algorithm (NKA) inspired by the mechanism of NK cells constructing phenotype detectors based on pathogen dose, which is a non-specific immune mechanism. NKA defines dose and phenotype detector, optimizes the detectors based on the memory evolution mechanism of phenotype, then establishes k-d tree for the optimized phenotype, and pathogens only need to match the dose with the nearest phenotype detector. Experimental results reveal that NKA can not only achieve the best performance through fewer detectors but also has a higher efficiency in the training and detection phase, compared with three versions of the distance-based NSA algorithms. Meanwhile, NKA generates comparable results compared with six popular machine learning algorithms.

关键词： Innate immune memory NK cell algorithm Artificial immune system Classification negative selection algorithm

来源：评论

学校读者我要写书评

暂无评论

A Framework of Intrusion Detection Inspired by Artificial Immune System 1st

A Framework of Intrusion Detection Inspired by Artificial Im...

引用

1st International Conference on Emerging Networking Architecture and Technologies (ICENAT)

作者： Zhang, Haodi Huang, Zeyuan Wang, Shuai Jin, Huamin Deng, Xiaodong China Telecom Corp Ltd Guangdong Res Inst Guangzhou Guangdong Peoples R China

ISBN: (纸本)9789811996962;9789811996979

The Human Immune System categorizes substances into self and non-self to find intrusions. This idea provides a good model for intrusion detection systems in the computer science field. This model has developed into a research field called Artificial Immune Systems. Such a system uses several algorithms and concepts like negative selection, a classical algorithm in this field. This paper takes an overview of the strengths and weaknesses of negative selection. It proposes the design of Artificial Immune System(AIS) based Intrusion Detection Systems (IDSs), which consists of three layers to attain accommodative immune response and an improved algorithm to decrease false positives and false negatives. Experiments on the same dataset prove the validity of our proposal and some specific indicator values can achieve the performance of the classical Artificial Intelligence(AI) algorithm.

关键词： Artificial Immune System negative selection algorithm Intrusion Detection

来源：评论

学校读者我要写书评

暂无评论

Using a Novel negative selection Inspired Anomaly Detection algorithm to Identify Corrupted Ribo-seq and RNA-seq Samples 19

Using a Novel Negative Selection Inspired Anomaly Detection ...

引用

10th ACM International Conference on Bioinformatics, Computational Biology and Health Informatics (ACM-BCB)

作者： Perkins, Patrick Heber, Steffen North Carolina State Univ Bioinformat Res Ctr Raleigh NC 27695 USA North Carolina State Univ Dept Comp Sci Raleigh NC 27695 USA

ISBN: (纸本)9781450366663

RNA-seq and Ribo-seq are popular techniques for quantifying cellular transcription and translation. These experiments use next-generation sequencing to produce genome-wide high-resolution snapshots of the total populations of mRNAs and translating ribosomes within the investigated samples. When performed in concert, these experiments yield valuable information about protein synthesis rates and translational efficiency. Due to their intricate experimental protocols and demanding data processing requirements, quality control and analysis of such experiments are often challenging. Therefore, methods for accurately assessing data quality, and for identifying contaminated samples, are greatly needed. In the following we use a novel negative selection inspired algorithm called Boundary Detection Using Nearest Neighbors (BDUNN), for the identification of corrupted samples. Our algorithm constructs a detector set and reduced training set that defines the boundaries between normal data points and potential anomalies. Subsequently, a nearest neighbor algorithm is used to classify unseen observations. We compare the performance of BDUNN with other popular negative selection and one-class classification algorithms, and show that BDUNN is capable of accurately and efficiently detecting anomalies in standard anomaly detection datasets and simulated RNA-seq and Ribo-seq data sets. Furthermore, we have implemented our method within an existing R Shiny platform for analyzing RNA-seq an Ribo-seq datasets, which permits downstream analysis of anomalous samples.

关键词： Ribosome profiling RNA-seq anomaly detection sample quality machine learning negative selection algorithm

来源：评论

学校读者我要写书评

暂无评论

Applying the negative selection algorithm for Merger and Acquisition Target Identification Theory and Case Study

Applying the Negative Selection Algorithm for Merger and Acq...

引用

1st BRICS Countries Congress on Computational Intelligence / 11th Brazilian Congress on Computational Intelligence (BRICS-CCI and CBIC)

作者： Paul, Satyakama Janacek, Andreas de Lima Neto, Fernando Buarque Marwala, Tshilidzi Univ Johannesburg Dept Mech Engn Sci FEBE APK Campus ZA-2006 Johannesburg South Africa Univ Johannesburg RTI ZA-2006 Johannesburg South Africa Univ Vienna Res Grp Theory & Applicat Algorithms A-1090 Vienna Austria Univ Permambuco POLI Comp Eng Program BR-720001 Recife PE Brazil

ISBN: (纸本)9781479931941

In this paper, we propose a new methodology based on the negative selection algorithm that belongs to the field of Computational Intelligence (specifically, Artificial Immune Systems - AIS) to identify takeover targets. Although considerable research based on customary statistical techniques and some contemporary Computational Intelligence techniques have been devoted to identify takeover targets, most of the existing studies are based upon multiple previous mergers and acquisitions. Contrary to previous research, the novelty of this proposal lies in the methodology's ability to suggest takeover targets for novice firms that are at the beginning of their merger and acquisition spree. We first discuss the theoretical perspective and then provide a case study with details for practical implementation, both capitalizing from unique generalization capabilities of AIS algorithms.

关键词： Merger and Acquisition M&A Takeover target prediction Artificial Immune System negative selection algorithm Euclidean distance Cosine similarity

来源：评论

学校读者我要写书评

暂无评论

Investigating One-Class Classifiers to Diagnose Alzheimer's Disease from Handwriting 21st

Investigating One-Class Classifiers to Diagnose Alzheimer's ...

引用

21st International Conference on Image Analysis and Processing (ICIAP)

作者： Parziale, Antonio Della Cioppa, Antonio Marcelli, Angelo Univ Salerno DIEM Via Giovanni Paolo II 132 I-84084 Fisciano SA Italy Univ Salerno Unit Natl Lab Artificial Intelligence & Intelligent Sy CINI Fisciano SA Italy ICAR CNR Via P Castellino 111 I-80131 Naples Italy

ISBN: (纸本)9783031064272;9783031064265

The analysis of handwriting and drawing has been adopted since the early studies to help diagnose neurodegenerative diseases, such as Alzheimer's and Parkinson's. Departing from the current state-of-the-art methods that approach the problem of discriminating between healthy subjects and patients by using two- or multi-class classifiers, we propose to adopt one-class classifier models, as they require only data by healthy subjects to build the classifier, thus avoiding to collect patient data, as requested by competing techniques. In this framework, we evaluated the performance of three models of one-class classifiers, namely the negative selection algorithm, the Isolation Forest and the One-Class Support Vector Machine, on the DARWIN dataset, which includes 174 subjects performing 25 handwriting/drawing tasks. The comparison with the state-of-the-art shows that the methods achieve state-of-the-art performance, and therefore may represent a viable alternative to the dominant approach.

关键词： Alzheimer's disease diagnosis Handwriting analysis One-class classifiers negative selection algorithm Isolation forest One-class support vector machine

来源：评论

学校读者我要写书评

暂无评论

Improving damage detection by combining multiple classifiers in different feature spaces

引用

ENGINEERING STRUCTURES 2024年 299卷

作者： Barontini, Alberto Masciotta, Maria Giovanna Amado-Mendes, Paulo Ramos, Luis F. Lourenco, Paulo B. Univ Minho ISISE ARISE Dept Civil Engn Guimaraes Portugal Univ G Annunzio Chieti Pescara Dept Engn & Geol Viale Pindaro 42 I-65127 Pescara Italy Univ G Annunzio Chieti Pescara Res Ctr UdA TechLab I-66100 Chieti Italy Univ Coimbra ISISE ARISE Dept Civil Engn Coimbra Portugal

Vibration-based damage identification methods have been explored in several engineering fields. However, the effective damage-sensitivity of vibration-based features and the balance between false alarms and readiness of the anomaly detection are still open issues. Here, a cost-effective damage detection methodology based on a Deterministically Generated negative selection algorithm is implemented and tested over a wide dataset generated through the numerical simulation of a three-storey concrete building in different damage scenarios. The approach allows training the classifiers only based on undamaged samples and ensures a prompt response to damage outbreak by fusing the classification conducted in parallel on different feature spaces, exploiting their distinct sensitivity to damage. A strategy to improve the reliability of the alarm is implemented by introducing counters and thresholds on consecutive classification outputs. A satisfactory trade-off between false and true alarms is achieved by comparing their probabilities, obtained through Markov chains, for different values of such thresholds.

关键词： Structural health monitoring Damage detection negative selection algorithm Damage sensitivity Multiple classifiers combination

来源：评论

学校读者我要写书评

暂无评论

Dealing with Complexity for Immune-Inspired Anomaly Detection in Cyber Physical Systems 9th

Dealing with Complexity for Immune-Inspired Anomaly Detectio...

引用

9th International Conference on Secure Knowledge Management in Artificial Intelligence Era (SKM)

作者： Reuter, Lenhard Leitner, Maria Smith, Paul Koschuch, Manuel Univ Appl Sci FH Campus Wien Vienna Austria Univ Vienna Fac Comp Sci Vienna Austria AIT Austrian Inst Technol GmbH Ctr Digital Safety & Secur Seibersdorf Austria

ISBN: (纸本)9783030975326;9783030975319

With digitization, critical infrastructures face a higher risk of security incidents and attacks on cyber-physical systems (CPS). In the past 50 years, research and practice have developed various approaches to monitor and detect attacks such as with anomaly detection. While many approaches focuses on artificial neural networks, bio-inspired approaches utilize nature as reference. For example, artificial immune systems (AIS) refer to principles of the natural immune system. In this paper, we investigate the negative selection algorithm (NSA), an algorithm from the domain of AIS for anomaly detection in CPS. Particularly in CPS, datasets can become quite complex and can require a number of detectors for the analysis. Therefore, we will investigate how AIS can be extended to handle and manage complex datasets of CPS. We propose two models that use Principal Component Analysis (PCA) and Autoencoder (AE) to enable dimensionality reduction. Using these models, we are able to show that it is possible to apply the NSA approach to such datasets. Our results indicate that the use of PCA and AE is beneficial for both a better representation of the data and therefore significantly relevant for an improvement of the detection rate, and provides in addition the possibility to add further features to support the identification of anomalies. As the NSA approach allows for distributed computation, it might be possible to allow faster or distributed detection;the extent to which this is possible remains to be investigated and therefore represents future work.

关键词： Anomaly detection negative selection algorithm Cyber physical system Principal component analysis Autoencoder

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：