ABC-FL: Anomalous and Benign client Classification in Federated Learning

Abstract

Federated Learning is a distributed machine learning framework designed for data privacy preservation i.e., local data remain private throughout the entire training and testing procedure. Federated Learning is gaining popularity because it allows one to use machine learning techniques while preserving privacy. However, it inherits the vulnerabilities and susceptibilities raised in deep learning techniques. For instance, Federated Learning is particularly vulnerable to data poisoning attacks that may deteriorate its performance and integrity due to its distributed nature and inaccessibility to the raw data. In addition, it is extremely difficult to correctly identify malicious clients due to the non-Independently and/or Identically Distributed (non-IID) data. The real-world data can be complex and diverse, making them hardly distinguishable from the malicious data without direct access to the raw data. Prior research has focused on detecting malicious clients while treating only the clients having IID data as benign. In this study, we propose a method that detects and classifies anomalous clients from benign clients when benign ones have non-IID data. Our proposed method leverages feature dimension reduction, dynamic clustering, and cosine similarity-based clipping. The experimental results validates that our proposed method not only classifies the malicious clients but also alleviates their negative influences from the entire procedure. Our findings may be used in future studies to effectively eliminate anomalous clients when building a model with diverse data. CCS CONCEPTS• Security and privacy; • Computing methodologies → Machine learning; Distributed artificial intelligence;

Open problems in medical federated learning

“…The aforementioned work's objective, anomaly detection, remains the same, but several approaches leveraged clustering-based and thresholding-based approaches (Shen et al, 2016;Fang et al, 2020;Tolpegin et al, 2020;Jeong et al, 2021;Cao et al, 2020;Sun et al, 2019). Auror dealt with targeted poisoning attacks leveraging clustering and thresholding techniques.…”

“…Their reweighting strategy used iteratively reweighted least squares to integrate repeated median regression. On the other hand, there are various defense attempts to prepare for possible attacks in the form of classifying anomalies (Li et al, 2020a;Shen et al, 2016;Fang et al, 2020;Tolpegin et al, 2020;Jeong et al, 2021). Li proposed spectral anomaly detection mechanism based on models' low-dimensional embeddings (Li et al, 2020a).…”

“…and Sun et al (2019),Jeong et al (2021) proposed anomalous and benign client classification in federated learning that leverages feature Medical federated learning dimension reduction, dynamic clustering and cosine-similarity-based clipping to detect and classify anomalous and benign clients where the benign ones have nonIID data and IID data. Similar to the aforementioned approaches, only the benign clients' model weights are aggregated in the global server.…”