Security and Privacy in Distributed Machine Learning

Dear Colleagues,

Decentralized machine learning involves training models on distributed data sources, often without directly sharing the raw data. As the field of machine learning expands and embraces decentralized architectures, ensuring the security of decentralized machine learning becomes crucial. Decentralized machine learning security focuses on developing innovative techniques, algorithms, and frameworks to guarantee the privacy, integrity, and confidentiality of decentralized machine learning systems. It involves developing mechanisms to prevent privacy leakage and unauthorized access to sensitive data during the training process. Ensuring the reliability and trustworthiness of the participants is crucial to prevent adversarial attacks or manipulation of the training process. Additionally, decentralized machine learning security involves addressing resource constraints, optimizing computation and communication overhead, and mitigating the risks associated with system vulnerabilities and attacks.

Distributed Machine Learning Security is an important research area that aims to address the security challenges arising from the distributed nature of machine learning systems. By developing robust privacy-preserving techniques, protecting the integrity of models, and securing communication infrastructure, researchers are working towards enabling the widespread adoption of distributed machine learning in various sensitive domains while ensuring data privacy and model security.

ICA3PP (established in 1995) is a famous, worldwide event that covers many dimensions of parallel algorithms and architectures, encompassing fundamental theoretical approaches, practical experimental projects, and commercial components and systems. The ICA3PP 2024 Workshop on Distributed Machine Learning Security, organized by the City University of Macau, is the 24th conference in this series. With the booming computing demands from every aspect of modern society, parallel processing has become increasingly critical and challenging. This conference provides a forum for academics and practitioners from all over the world to exchange ideas on improving the efficiency, performance, reliability, security, and interoperability of computing systems and applications.

The Special Issue primarily represents a collection of extended versions of selected papers presented at the ICA3PP 2024 Workshop on Distributed Machine Learning Security. However, papers not presented at the ICA3PP are also welcome. The topics of interest include, but are not limited to, the following:

• Privacy-preserving techniques in decentralized machine learning;
• Secure multi-party computation for distributed machine learning;
• Federated learning;
• Detection and mitigation of model poisoning attacks in decentralized settings;
• Secure communication protocols for decentralized machine learning;
• Trustworthiness and reputation management in decentralized machine learning;
• Anomaly detection and intrusion detection in distributed machine learning;
• Resource-constrained decentralized machine learning security;
• Scalability and efficiency of security mechanisms in decentralized machine learning;
• Secure aggregation methods for distributed machine learning;
• Cryptographic protocols for secure data sharing in decentralized settings;
• Adversarial attacks and defenses in decentralized machine learning;
• Standardization and interoperability in decentralized machine learning security;
• Real-world applications and case studies of decentralized machine learning security.

Dr. Chuan Zhang
Dr. Xiangyun Tang
Dr. Yajie Wang
Guest Editors

Manuscript Submission Information

Manuscripts should be submitted online at www.mdpi.com by registering and logging in to this website. Once you are registered, click here to go to the submission form. Manuscripts can be submitted until the deadline. All submissions that pass pre-check are peer-reviewed. Accepted papers will be published continuously in the journal (as soon as accepted) and will be listed together on the special issue website. Research articles, review articles as well as short communications are invited. For planned papers, a title and short abstract (about 100 words) can be sent to the Editorial Office for announcement on this website.

Submitted manuscripts should not have been published previously, nor be under consideration for publication elsewhere (except conference proceedings papers). All manuscripts are thoroughly refereed through a single-blind peer-review process. A guide for authors and other relevant information for submission of manuscripts is available on the Instructions for Authors page. Electronics is an international peer-reviewed open access semimonthly journal published by MDPI.

Please visit the Instructions for Authors page before submitting a manuscript. The Article Processing Charge (APC) for publication in this open access journal is 2400 CHF (Swiss Francs). Submitted papers should be well formatted and use good English. Authors may use MDPI's English editing service prior to publication or during author revisions.

• privacy-preserving techniques
• secure multi-party computation
• federated learning
• poisoning attacks
• communication protocols
• secure aggregation methods
• artificial intelligence security