Learning Graph Similarity With Large Spectral Gap

Wu, Zongze; Liu, Sihui; Ding, Chris; Ren, Zhigang; Xie, Shengli

doi:10.1109/tsmc.2019.2899398

Cited by 16 publications

(7 citation statements)

References 53 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…K-Means and spectral clustering with normalized cut (Ncut) are chosen for the baseline. In addition, classical sparse subspace clustering (SSC) [21] and four related algorithms in recent years are adopted for further analysis, including RSEC [22], FastESC [23], USENC [24] and JSPC [25]. We choose four general performance indicators to evaluate the effect, including accuracy (ACC), normalized mutual information (NMI), adjusted rand index (ARI) and F-measure (F).…”

Section: Comparison Methodsmentioning

confidence: 99%

Spectral clustering based on high‐frequency texture components for face datasets

Liang

Guo

Liu

et al. 2021

IET Image Processing

View full text Add to dashboard Cite

Spectral clustering is one of the most widely used technologies for clustering tasks, which represents data as a weighted graph, and aims to find an appropriate way to cut the graph apart in order to categorize the raw data. The pivotal step of spectral clustering is to find out the accurate information to estimate the relationship of pairwise data, based on which a graph can be constructed. According to the cognition that different faces are distinguished by the edge contour which can be represented by high-frequency texture components, a novel spectral clustering algorithm via high-frequency signal of human face, named highfrequency spectral clustering (HFSC) is proposed. In HFSC, first the local high-frequency texture components are extracted from samples. Then the relationship of pairwise samples can be estimated with the degree of correlation, which is produced from the image high-frequency information. The graph will be set up with the correlation information. Subsequently the graph cut will be implemented to achieve the final clustering results. Experimental results show that this algorithm outperforms the state-of-the-art clustering methods on several datasets.This is an open access article under the terms of the Creative Commons Attribution License, which permits use, distribution and reproduction in any medium, provided the original work is properly cited.

show abstract

Section: Comparison Methodsmentioning

confidence: 99%

Spectral clustering based on high‐frequency texture components for face datasets

Liang

Guo

Liu

et al. 2021

IET Image Processing

View full text Add to dashboard Cite

show abstract

“…Machine learning-based IDPS are particularly effective in evolving threat landscapes, as they can learn from new data and adapt their detection strategies accordingly, enhancing their ability to identify and mitigate threats before they can cause significant harm. This proactive approach not only improves the accuracy of threat detection but also reduces the response time, allowing organizations to address vulnerabilities and mitigate attacks more swiftly and effectively (Wu et al, 2021). By integrating advanced IDPS into their cybersecurity framework, organizations can significantly enhance their defense mechanisms, ensuring a more secure and resilient network environment.…”

Section: Intrusion Detection and Prevention Systems (Idps)mentioning

confidence: 99%

“…Data Loss Prevention (DLP) solutions are crucial for safeguarding sensitive information by preventing unauthorized data exfiltration through continuous monitoring and control of data transfers. These solutions are designed to detect, monitor, and block the unauthorized transmission of sensitive data, thereby mitigating the risk of data breaches (Wu et al, 2021). Content-aware DLP systems enhance this capability by analyzing data content in real-time to identify sensitive information, such as personal identifiable information (PII), financial records, or intellectual property.…”

Section: Data Loss Prevention (Dlp)mentioning

confidence: 99%

Advanced Cybersecurity Protocols for Securing Data Management Systems in Industrial and Healthcare Environments

Joy,

Islam,

Rahaman

et al. 2024

GMJ

View full text Add to dashboard Cite

The increasing digitalization of industrial and healthcare environments has heightened the importance of cybersecurity to protect sensitive data and ensure the integrity of data management systems. This article examines advanced cybersecurity protocols designed to address the unique challenges faced by these sectors. It explores various threats, including ransomware, phishing, and insider threats, as well as system vulnerabilities found in legacy systems, IoT and medical devices, and cloud computing. The implementation of robust security measures such as encryption, multi-factor authentication, and intrusion detection systems is analyzed in depth. Through a comprehensive review of existing research and real-world case studies, including those of Maersk Line, Memorial Health System, and Norsk Hydro, this study provides insights into effective strategies for enhancing cybersecurity in industrial and healthcare settings. The findings highlight the critical role of these measures, alongside regular staff training and awareness programs, in safeguarding data management systems. By leveraging advanced technologies and proactive security practices, organizations can better protect their sensitive data against evolving cyber threats.

show abstract

“…Another well-known theory is spectral analysis, which builds a graph with data points as nodes and similarities between points as edges, and then studies the eigenvalues of the normalized graph Laplacian matrix [29]. The graph cut theory suggests that small eigenvalues of the Laplacian matrix reflect weak connections of components, and a large eigen gap indicates a proper cluster number [30]. The advantages of spectral clustering include the detection of various cluster shapes, utilization of kernel functions, and cluster number estimation [29].…”

Section: B Automatic Clustering Algorithmsmentioning

confidence: 99%

An Integrated Cluster Detection, Optimization, and Interpretation Approach for Financial Data

Tie

Kou

et al. 2022

IEEE Trans. Cybern.

144

View full text Add to dashboard Cite

In many financial applications, such as fraud detection, reject inference, and credit evaluation, detecting clusters automatically is critical because it helps to understand the subpatterns of the data that can be used to infer user's behaviors and identify potential risks. Due to the complexity of human behaviors and changing social environments, the distributions of financial data are usually complex and it is challenging to find clusters and give reasonable interpretations. The goal of this study is to develop an integrated approach to detect clusters in financial data, and optimize the scope of the clusters such that the clusters can be easily interpreted. Specifically, we first proposed a new cluster quality evaluation criterion, which is free from large-scale computation and can guide base clustering algorithms such as k-Means to detect hyperellipsoidal clusters adaptively. Then, we designed a new solver for a revised support vector data description model, which efficiently refines the centroids and scopes of the detected clusters to make the clusters tighter such that the data in the clusters share greater similarities, and thus, the clusters can be easily interpreted with eigenvectors. Using ten financial datasets, the experiments showed that the proposed algorithm can efficiently find reasonable number of clusters. The proposed approach is suitable for large-scale financial datasets whose features are meaningful, and also applicable to financial mining tasks, such as data distribution interpretation and anomaly detection. Index Terms-Clustering methods, data mining, financial management, spectral analysis. I. INTRODUCTIONI N MANY financial applications, such as credit evaluation, fraud detection, and reject inference, labeled data of ground truth are highly scarce. Hence, unsupervised models are used intensively to infer the patterns behind the data. Compared to supervised learning, unsupervised learning is more challenging because of lacking objective criteria to guide the learning Manuscript

show abstract

Learning Graph Similarity With Large Spectral Gap

Cited by 16 publications

References 53 publications

Spectral clustering based on high‐frequency texture components for face datasets

Spectral clustering based on high‐frequency texture components for face datasets

Advanced Cybersecurity Protocols for Securing Data Management Systems in Industrial and Healthcare Environments

An Integrated Cluster Detection, Optimization, and Interpretation Approach for Financial Data

Contact Info

Product

Resources

About