Nothing Special   »   [go: up one dir, main page]

×
Please click here if you are not redirected within a few seconds.
In this work, we improve our previous research, score-based aggregation (SBA), and develop a malicious-client elimination algorithm before aggregation.
Feb 15, 2023 · In this work, we improve our previous research, score-based aggregation (SBA), and develop a malicious-client elimination algorithm before aggregation.
Oct 1, 2023 · How to cope with malicious federated learning clients: : An unsupervised learning-based approach. Authors: Murat Arda Onsu.
Onsu et al. (2023) proposed an unsupervised FL approach to identify malicious clients using a Score-Based Aggregation (SBA) that showed significant performance ...
Coping. Preprint. How to Cope with Malicious Federated Learning Clients: an Unsupervised Learning-Based Approach. January 2023. DOI:10.2139/ssrn.4359318.
People also ask
How to cope with malicious federated learning clients: An unsupervised learning-based approach ... training method is proposed, namely federated learning ...
Aug 1, 2024 · Our approach significantly outperforms random client selection methods, achieving approximately 60% accuracy in the presence of 10% malicious clients.
Aug 16, 2024 · We propose a simple yet effective framework to detect malicious clients, namely Confidence-Aware Defense (CAD), that utilizes the confidence scores of local ...
Missing: cope unsupervised
Mar 18, 2024 · This makes the system vulnerable to malicious attacks when dealing with dishonest clients. In this work, we address this problem by ...
Aug 5, 2024 · However, FL systems are vulnerable to attacks that are happening in malicious clients through data poisoning and model poisoning, which can ...
Missing: based | Show results with:based