Latest Breaking News On - Federated learning - Page 1 : comparemela.com

Innovative Data Management: Tech Leaders Share Leading-Edge Approaches

For businesses, getting the biggest ROI from collected data requires strategic, updated data management strategies.

INPHER's Enterprise-Ready SecurAI Performantly Protects LLMs with NVIDIA Confidential Computing

Connected Car News: Infineon, Seyond, Peachtree Corners, Diodes, WindRiver, DENSO, Seeing Machines, Maganachip, eSync Alliance & Infineon.

"FLPurifier: Backdoor Defense in Federated Learning vi" by Jiale Zhang, Chengcheng Zhu et al.

Recent studies have demonstrated that backdoor attacks can cause a significant security threat to federated learning. Existing defense methods mainly focus on detecting or eliminating the backdoor patterns after the model is backdoored. However, these methods either cause model performance degradation or heavily rely on impractical assumptions, such as labeled clean data, which exhibit limited effectiveness in federated learning. To this end, we propose FLPurifier, a novel backdoor defense method in federated learning that can effectively purify the possible backdoor attributes before federated aggregation. Specifically, FLPurifier splits a complete model into a feature extractor and classifier, in which the extractor is trained in a decoupled contrastive manner to break the strong correlation between trigger features and the target label. Compared with existing backdoor mitigation methods, FLPurifier doesn’t rely on impractical assumptions since it can effectively purify the backdoo

"A Review of Solving Non-IID Data in Federated Learning: Current Status" by Wenhai Lu, Jieren Cheng et al.

Federated learning (FL), as a machine learning framework, has garnered substantial attention from researchers in recent years. FL makes it possible to train a global model through coordination by a central server while ensuring the privacy of data on individual edge devices. However, the data on edge devices that participate in FL training are not independently and identically distributed (IID), resulting in challenges related to heterogeneity data. In this paper, we introduce the challenges generated by non-IID data to FL and provide a detailed classification of non-IID data. Then, we summarize the existing solutions to non-IID data in FL from the perspectives of data and process. To the best of our knowledge, despite the considerable efforts achieved by many researchers in solving the non-IID problem, some issues remain unsolved. This paper provides researchers with the latest findings and analyzes the potential future directions for solving non-IID in FL.