Mediaspace scheduled maintenance: Aug 25, 2026 07:00 - 12:00 AM. During this time, videos will be temporarily unavailable. Check status updates.

Lecture

Decentralized Learning Algorithms: A Systems View

Description

This lecture explores the challenges of Federated Learning (FL) and Decentralized Learning (DL) systems, focusing on issues like expensive communication and data center problems. The instructor discusses strategies to enhance these systems through decentralization, data heterogeneity, and privacy guarantees, presenting a TEE-based decentralized recommender system as a case study. The lecture also introduces the Rex++ extension for ML tasks and the GeL algorithm for boosting federated learning. Overall, the lecture provides insights into improving the scalability, privacy, and efficiency of decentralized learning algorithms.

Official source

Related concepts (25)

Machine learning

Machine learning (ML) is an umbrella term for solving problems for which development of algorithms by human programmers would be cost-prohibitive, and instead the problems are solved by helping machines 'discover' their 'own' algorithms, without needing to be explicitly told what to do by any human-developed algorithms. Recently, generative artificial neural networks have been able to surpass results of many previous approaches.

Privacy law

Privacy law is the body of law that deals with the regulating, storing, and using of personally identifiable information, personal healthcare information, and financial information of individuals, which can be collected by governments, public or private organisations, or other individuals. It also applies in the commercial sector to things like trade secrets and the liability that directors, officers, and employees have when handing sensitive information.

Privacy

Privacy (UK, US) is the ability of an individual or group to seclude themselves or information about themselves, and thereby express themselves selectively. The domain of privacy partially overlaps with security, which can include the concepts of appropriate use and protection of information. Privacy may also take the form of bodily integrity. There have been many different conceptions of privacy throughout history. Most cultures recognize the right of an individual to withhold aspects of their personal lives from public record.

Right to privacy

The right to privacy is an element of various legal traditions that intends to restrain governmental and private actions that threaten the privacy of individuals. Over 150 national constitutions mention the right to privacy. On 10 December 1948, the United Nations General Assembly adopted the Universal Declaration of Human Rights (UDHR), originally written to guarantee individual rights of everyone everywhere; while right to privacy does not appear in the document, many interpret this through Article 12, which states: "No one shall be subjected to arbitrary interference with his privacy, family, home or correspondence, nor to attacks upon his honour and reputation.

Communication

Communication is usually defined as the transmission of information. The term can also refer to the message itself, or the field of inquiry studying these transmissions, also known as communication studies. The precise definition of communication is disputed. Controversial issues are whether unintentional or failed transmissions are included and whether communication does not just transmit meaning but also create it. Models of communication aim to provide a simplified overview of its main components and their interaction.

Related lectures (31)

Machine Learning Security

Explores machine learning security, including model stealing, altering outputs, adversarial conditions, and privacy challenges, emphasizing the importance of addressing biases in machine learning models.

Data Privacy in eHealth Research

Explores data privacy challenges and perspectives in eHealth research, focusing on GDPR compliance, sensitive health data management, and decentralized agents.

Privacy: Threat Modeling and Differential Privacy

Explores the significance of privacy, risks of targeted advertising, and the concept of differential privacy.

Legal Aspects: Data Protection and Privacy

Explores Swiss legal framework on data protection, privacy, cybercrimes, and the role of the Federal Data Protection Commissioner.

Trust and Privacy