检索结果-内蒙古大学图书馆

Securing templates in a face recognition system using error-correcting output code and chaos theoryy

COMPUTERS & ELECTRICAL ENGINEERING 2018年 72卷 644-659页

作者： Nazari, Sara Moin, Mohammad-Shahram Kanan, Hamidreza Rashidy Islamic Azad Univ Qazvin Branch Fac Comp & Informat Technol Engn Qazvin Iran ICT Res Inst Fac Informat Technol Tehran Iran Shahid Rajaee Teacher Training Univ Dept Comp Engn Tehran Iran

In biometric cryptosystems, biometric data is combined with cryptography algorithms to generate secure templates. In these systems, creating protected templates with both high discriminability and high security is a challenging issue. To address this issue, this paper proposes a new face cryptosystem based on binarization transformation, chaos feature permutation and fuzzy commitment scheme. To enhance discriminability, real-valued templates are converted into their binary versions using a new discriminant binarization transformation based on error-correcting output code. Then, the chaos feature permutation is used to increase the security and privacy of binary templates, and also to protect the fuzzy commitment scheme against cross-matching attacks. The proposed scheme is evaluated on three well-known face databases, i.e. CMU PIE, FEI, and Extended Yale B. Experimental results show that the proposed method improves discriminability, as well as privacy and security of the system, compared to the existing face template protection algorithms. (C) 2018 Elsevier Ltd. All rights reserved.

关键词： Face template protection Cryptosystem Chaos feature permutation error-correcting output code Discriminant binarization transformation

来源：评论

学校读者我要写书评

暂无评论

Efficient error-correcting output codes for Adversarial Learning Robustness

Efficient Error-correcting Output Codes for Adversarial Lear...

引用

IEEE International Conference on Communications (ICC)

作者： Wan, Li Alpcan, Tansu Viterbo, Emanuele Kuijper, Margreta Univ Melbourne Dept Elect & Elect Engn Parkville Vic 3010 Australia Monash Univ Dept Elect & Comp Syst Engn Clayton Vic 3800 Australia

ISBN: (数字)9781538683477

ISBN: (纸本)9781538683477

Despite their many successful applications, Deep Neural Networks (DNNs) are vulnerable to intentionally designed adversarial examples. Adversarial robustness describes the ability of a machine learning model, e.g., a neural network, to defend against such adversarial attacks. In coding theory, codebooks are designed to minimize the impact of errors occurring with transmission through a noisy channel. Motivated by the similarities between passing a codeword through a noisy channel and defending against adversarial attacks, error-correcting output codes (ECOCs) are used to achieve state-of-the-art adversarial robustness. Research on codebook designs and the association of codewords to classification labels (assignment) is still at the very early stages, with great room for improvement. In this work, we present novel codebook design and assignment procedures in two stages due to the complexity (NP-hardness) of the underlying problem. A rule-based heuristic codebook design method is proposed in the first stage and an optimization problem to assign the codewords to labels is proposed in the second stage. Since this optimization is NP-hard, a greedy algorithm is proposed to provide a sub-optimal solution. We demonstrate the effectiveness of our framework on three benchmark datasets, under different types of adversarial attacks. The experimental results show that our error-correcting output code framework can effectively improve the adversarial robustness of machine learning models, with up to a 10% increase in accuracy.

关键词： error-correcting output code machine learning adversarial robustness coding theory

来源：评论

学校读者我要写书评

暂无评论

A self-adaptive soft-recoding strategy for performance improvement of error-correcting output codes

引用

PATTERN RECOGNITION 2023年第1期143卷

作者： Lin, Guangyi Gao, Jie Zeng, Nan Xu, Yong Liu, Kunhong Wang, Beizhan Yao, Junfeng Wu, Qingqiang Xiamen Univ Sch Informat Xiamen 361005 Peoples R China Xiamen Univ Technol Sch Elect Engn & Automat Xiamen 361024 Peoples R China Fujian Med Univ Sch Publ Hlth Fuzhou 350122 Peoples R China Fujian Univ Technol Sch Comp Sci & Math Fuzhou 350016 Peoples R China

The technique of error-correcting output codes (ECOC) has been proven to be of high discriminative ability in many classification applications. However, most algorithms on the ECOC were designed based on the binary or ternary codes (referred to as the hard codes), which might fail to precisely correct errors in dealing with tough tasks. In this study, a Soft-Recoding strategy based on a self-adaptive algorithm is proposed, which replaces the traditional hard codes with the real-value elements to better fit the output distribution of the base learners. This is achieved by minimizing the ratio of two distances: the distance of the output vector to the ground-truth class, and the average distance of the output vector to the remaining classes. Extensive experiments using five different hard ECOC algorithms and the corresponding softened versions on twenty datasets with diversified numbers of features and classes confirm the effectiveness of our Soft-Recoding strategy in promoting the performance of the original ECOC algorithms. Our source code and additional results are available at: ***/MLDMXM2017/ SA-soft-recoding.

关键词： error-correcting output code Soft-recoding Ternary coding Regression Multiclass classification

来源：评论

学校读者我要写书评

暂无评论

A Novel Hyper-Tuned Multilayer Perceptron With Effective Stochastic Learning Strategies for Missing Values Imputation

引用

EXPERT SYSTEMS 2025年第2期42卷

作者： Siddiqi, Muhammad Hameed Alruwaili, Madallah Alhwaiti, Yousef Alanazi, Saad Khan, Faheem Jouf Univ Coll Comp & Informat Sci Sakaka Aljouf Saudi Arabia Gachon Univ Dept Comp Sci Seongnam SI South Korea

A vast amount of data in many different formats is produced and stored daily, offering machine learning a valuable resource to enhance its predictive capabilities. However, the pervasiveness of inaccuracies in real-world data presents a significant barrier that can seriously limit the effectiveness of learning algorithms. The ensemble models and hyper-tuned multi-layer perceptron (MLP) with need-based hidden neuron layers are effective frameworks for data imputation. Addressing the issue of missing data is a complex and demanding task, and much remains to be explored in developing effective and precise methods for predicting and imputing missing values across different datasets. The study offers important perspectives on using algorithms in machine learning to predict and impute gaps in data in recently updated datasets. The findings indicate that finely tuned MLP classifiers notably improve prediction accuracy and dependability compared to models with a static or reduced number of neurons. Furthermore, the study highlights the promising potential of ensemble models within the error-correcting output code (ECOC) framework as an effective approach for this task. It also suggests future research directions to refine further and strengthen machine learning-based imputation methods regarding precision and stability. ECOC framework includes all kinds of MLP classifiers and regressors such as binary classifiers, multi-class classifiers, or regression models. MLP models predict complex relationships in modern datasets. Hugging Face, COSMIC, SKlearn, and Kaggle have relevant and updated datasets. The weighted average recognition (96%) shows that the proposed MLP-based stochastic learning strategies achieved better results.

关键词： data imputation deep learning error-correcting output code missing values multi-layer perceptron multiple imputation

来源：评论

学校读者我要写书评

暂无评论

Multiclass False Data Injection Attacks Detection and Classification in Automatic Generation Control

Multiclass False Data Injection Attacks Detection and Classi...

引用

IEEE Canadian Conference on Electrical and Computer Engineering (CCECE)

作者： Alshareef, Sami M. Jouf Univ Fac Engn Elect Engn Dept Sakaka Aijouf Saudi Arabia

ISBN: (纸本)9798350371635;9798350371628

The reliance of automatic generation control (AGC) on modern communication exposes it to potential cyber-attacks, particularly false data injection (FDI) attacks, posing significant risks to power systems. While existing research has addressed the impact and detection of these attacks, a gap remains in their effective classification. To bridge this gap, this paper introduces an innovative machine learning approach to accurately classify four FDI attack types (pulse, random, ramp, and scaling). Utilizing original signal data from the AGC power system, measuring frequency and tie line power deviations, the research employs error-correcting output codes with a sparse random coding design for multiclass classification. Three binary learners, namely support vector machine (SVM), k-nearest neighbors (KNN), and decision tree (DT), are employed to classify these FDI attacks. The simulation results highlight the DT binary learner's performance achieving higher precision, recall, and F-score compared to KNN and SVM. This precise multiclass FDI classification establishes a solid foundation for proposing effective mitigation strategies and enhancing cybersecurity measures.

关键词： False data injection automatic generation control error-correcting output code multiclass classification machine learning

来源：评论

学校读者我要写书评

暂无评论

A novel error-correcting output codes based on genetic programming and ternary digit operators

引用

PATTERN RECOGNITION 2021年 110卷 107642-107642页

作者： Yi-Fan, Liang Chang, Liu Han-Rui, Wang Kun-Hong, Liu Jun-Feng, Yao Ying-Ying, She Gui-Ming, Dai Okina, Yuna Xiamen Univ Sch Informat Xiamen 361005 Fujian Peoples R China Fudan Univ Sch Comp Sci Shanghai Key Lab Intelligent Informat Proc Shanghai Peoples R China Sumitomo Elect Ind Ltd R&D Unit Frontier Technol Lab 1 Taya Cho Yokohama Kanagawa 2448588 Japan

The key to the success of an error-correcting output code (ECOC) algorithm is the effective codematrix, which represents a set of class reassignment schemes for decomposing a multiclass problem into a set of binary class problems. This paper proposes a new method, which uses Ternary digit Operators based Genetic Programming (GP) to generate effective ECOC codematrix (TOGP-ECOC for short). In our GP, each terminal node stores a ternary digit string, representing a column and a related feature subset;each non-terminal node represents a ternary digit operator, which produces a new column based on its child nodes. In this way, each individual is interpreted as an ECOC codematrix along with a set of corresponding feature subsets, serving the solution for the multiclass classification task. When a new individual is produced, a legality checking process is carried out to verify whether the transformed codematrix follows the ECOC constraints. The illegal one is corrected according to different strategies. Besides, a local optimization algorithm is designed to prune redundant columns and improve the performance of each individual. Our experiments compared TOGP-ECOC with some well known ECOC algorithms on various data sets, and the results confirm the superiority of our algorithm. Our source code is available at: https://***/MLDMXM2017/TOGP-ECOC. (c) 2020 Elsevier Ltd. All rights reserved.

关键词： error-correcting output code Ternary digit operator Genetic programming Feature selection

来源：评论

学校读者我要写书评

暂无评论

error-correcting output codes for multi-label emotion classification

引用

MULTIMEDIA TOOLS AND APPLICATIONS 2016年第22期75卷 14399-14416页

作者： Li, Chao Feng, Zhiyong Xu, Chao Tianjin Univ Sch Comp Sci & Technol Tianjin Peoples R China Tianjin Univ Sch Comp Software Tianjin Peoples R China

Multi-modal affective data such as EEG and physiological signals is increasingly utilized to analyze of human emotional states. Due to the noise existed in collected affective data, however, the performance of emotion recognition is still not satisfied. In fact, the issue of emotion recognition can be regarded as channel coding, which focuses on reliable communication through noise channels. Using affective data and its label, the redundant codeword would be generated to correct signals noise and recover emotional label information. Therefore, we utilize multi-label output codes method to improve accuracy and robustness of multi-dimensional emotion recognition by training a redundant codeword model, which is the idea of error-correcting output codes. The experiment results on DEAP dataset show that the multi-label output codes method outperforms other traditional machine learning or pattern recognition methods for the prediction of emotional multi-labels.

关键词： Affective computing Multi-modal error-correcting output code EEG signals Physiological signals

来源：评论

学校读者我要写书评

暂无评论

Novel Hybrid Model for Music Genre Classification based on Support Vector Machine

Novel Hybrid Model for Music Genre Classification based on S...

引用

IEEE Symposium on Computer Applications and Industrial Electronics (ISCAIE)

作者： Sharma, Srishti Fulzele, Prasenjeet Sreedevi, Indu Delhi Technol Univ Dept Elect Engn New Delhi India Jaypee Inst Informat Technol Dept Comp Sci Noida Uttar Pradesh India Delhi Technol Univ Dept Elect & Commun Engn New Delhi India

ISBN: (纸本)9781538635278

Automatic Music genre classification using machine learning techniques has gathered momentum in the recent past and continues to govern this area of research with promising results, saving time and human efforts. Machine Learning techniques have successfully dominated the methods used for the automation of this classification. In this paper, a novel method of music genre classification is proposed based on stacking of Support Vector Machine (SVM) with Relevance Vector Machine (RVM) and Decision Trees. The model uses the acoustic properties of the audio files as their features for classification. The three models are trained as error-correcting output code Classifiers and they individually classify the audio files with certain posterior probabilities. The results from the three classifiers are fused using the sum rule to evaluate the final outcome of the combined model. The performance of this hybrid model is evaluated on the GTZAN music dataset and is compared with individual performances of the models used in the combination. The proposed combined model outperformed the other models with an accuracy of 87% confirming the efficient utilization of the advantages of individual models.

关键词： Music Genre Classification Machine Learning Support Vector Machine Relevance Vector Machine Decision Trees error-correcting output code

来源：评论

学校读者我要写书评

暂无评论

Multiclass classification of FDG PET scans for the distinction between Parkinson's disease and atypical parkinsonian syndromes

引用

NEUROIMAGE-CLINICAL 2013年第1期2卷 883-893页

作者： Garraux, Gatan Phillips, Christophe Schrouff, Jessica Kreisler, Alexandre Lemaire, Christian Degueldre, Christian Delcour, Christian Hustinx, Roland Luxen, Andre Destee, Alain Salmon, Eric Univ Liege Cyclotron Res Ctr B-4000 Liege Belgium Univ Liege Dept Elect Engn & Comp Sci B-4000 Liege Belgium Univ Hosp Ctr Dept Neurol B-4000 Liege Belgium Univ Hosp Ctr Dept Nucl Med B-4000 Liege Belgium Ctr Hosp Peltzer La Tourelle Dept Neurol B-4800 Verviers Belgium Univ Lille Nord France F-59000 Lille France UDSL Lille France CHU Lille Lille France

Most available pattern recognition methods in neuroimaging address binary classification problems. Here, we used relevance vector machine (RVM) in combination with booststrap resampling ('bagging') for non-hierarchical multiclass classification. The method was tested on 120 cerebral (18)fluorodeoxyglucose (FDG) positron emission tomography (PET) scans performed in patients who exhibited parkinsonian clinical features for 3.5 years on average but that were outside the prevailing perception for Parkinson's disease (PD). A radiological diagnosis of PD was suggested for 30 patients at the time of PET imaging. However, at follow-up several years after PET imaging, 42 of them finally received a clinical diagnosis of PD. The remaining 78 APS patients were diagnosed with multiple system atrophy (MSA, N = 31), progressive supranuclear palsy (PSP, N = 26) and corticobasal syndrome (CBS, N = 21), respectively. With respect to this standard of truth, classification sensitivity, specificity, positive and negative predictive values for PD were 93% 83% 75% and 96%, respectively using binary RVM (PD vs. APS) and 90%, 87%, 79% and 94%, respectively, using multiclass RVM (PD vs. MSA vs. PSP vs. CBS). Multiclass RVM achieved 45%, 55% and 62% classification accuracy for, MSA, PSP and CBS, respectively. Finally, a majority confidence ratio was computed for each scan on the basis of class pairs that were the most frequently assigned by RVM. Altogether, the results suggest that automatic multiclass RVM classification of FDG PET scans achieves adequate performance for the early differentiation between PD and APS on the basis of cerebral FDG uptake patterns when the clinical diagnosis is felt uncertain. This approach cannot be recommended yet as an aid for distinction between the three APS classes under consideration. (C) 2013 The Authors. Published by Elsevier Inc.

关键词： Computer-aided diagnosis Data mining Pattern recognition Boostrap resampling Bagging error-correcting output code Multiclass classification Relevance vector machine FDG PET Parkinson's disease Multiple system atrophy Progressive supranuclear palsy Corticobasal syndrome

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：