Robust Federated Opportunistic Learning in the Presence of Label Quality Disparity

Chengxi Li, Gang Li, Pramod K. Varshney

Research output: Contribution to journalArticlepeer-review

Abstract

In this paper, the problem of federated learning (FL) in the presence of label quality disparity is considered. To address this problem, Federated Opportunistic Computing for Ubiquitous System (FOCUS) has been proposed very recently. In FOCUS, the central server utilizes its accurately labeled benchmark samples to quantify the credibility of different clients by computing the cross-entropy(CE) loss of the locally updated models on the benchmark dataset and the CE loss of the global model on the local datasets. However, FOCUS assumes the availability of the accurate labels of the benchmark dataset, which is difficult to guarantee under many practical scenarios. To overcome this limitation of FOCUS, we propose a new algorithm named Robust Federated Opportunistic Learning (RFOL), which does not require the benchmark samples at the central server to be labeled. In RFOL, the client credibility is evaluated by computing the Kullback-Leibler (KL) divergence among the soft predictions on the benchmark samples of different locally updated models and the CE loss of the global model on the local datasets. The experimental results on several popular datasets reveal that, 1) with an unlabeled benchmark dataset at the server, the proposed RFOL algorithm attains almost the same learning performance as FOCUS which requires an accurately labeled benchmark dataset at the server; 2) with an inaccurately labeled benchmark dataset, RFOL outperforms FOCUS, which shows that the former is more robust to the inaccurate labels of the benchmark samples; 3) RFOL outperforms FedAvg which assigns equal credibility to all the clients.

Original languageEnglish (US)
JournalIEEE Internet of Things Journal
DOIs
StateAccepted/In press - 2021

Keywords

  • Benchmark testing
  • Computational modeling
  • Data models
  • Federated learning
  • Internet of Things
  • Kullback-Leibler (KL) divergence.
  • label quality
  • Predictive models
  • Servers
  • Training data

ASJC Scopus subject areas

  • Signal Processing
  • Information Systems
  • Hardware and Architecture
  • Computer Science Applications
  • Computer Networks and Communications

Fingerprint

Dive into the research topics of 'Robust Federated Opportunistic Learning in the Presence of Label Quality Disparity'. Together they form a unique fingerprint.

Cite this