FedES: Federated Early-Stopping for Hindering Memorizing Heterogeneous Label Noise

Authors: Bixiao Zeng, Xiaodong Yang, Yiqiang Chen, Zhiqi Shen, Hanchao Yu, Yingwei Zhang

IJCAI 2024 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental Extensive experiments conducted on varying synthetic and real-world label noise demonstrate the superior performance of Fed ES over the state-of-the-art methods.
Researcher Affiliation Academia 1Beijing Key Laboratory of Mobile Computing and Pervasive Device, Institute of Computing Technology, Chinese Academy of Sciences 2University of Chinese Academy of Sciences 3Peng Cheng Laboratory 4Nanyang Technological University 5Bureau of Frontier Sciences and Education, Chinese Academy of Sciences
Pseudocode Yes Algorithm 1: Fed ES: Federated Early-Stopping
Open Source Code No The paper does not provide concrete access to source code for the methodology described.
Open Datasets Yes 1. CIFAR-10/100 datasets [Krizhevsky et al., 2009]: ... Both datasets are divided into train/test sets with a ratio of 5:1. 2. Clothing1M dataset [Xiao et al., 2015]: ... Clothing1M is divided into train/test sets with a ratio of 4:1.
Dataset Splits No The paper mentions 'train/test sets' and '20% of the test set is reserved as a benchmark dataset so that some comparison methods can use it as a reference to evaluate noisy clients'. It does not specify a distinct validation split for its own model training or hyperparameter tuning.
Hardware Specification No The paper does not provide specific hardware details (exact GPU/CPU models, processor types with speeds, memory amounts, or detailed computer specifications) used for running its experiments.
Software Dependencies No The paper mentions 'The Res Net-18 [...] model serves as the backbone' and 'optimizer of SGD', but it does not provide specific ancillary software details with version numbers (e.g., library or solver names with version numbers like Python 3.8, PyTorch 1.9).
Experiment Setup Yes Implementation Details The Res Net-18 [He et al., 2016] model serves as the backbone for both the CIFAR-10/100 and Clothing1M datasets. For the CIFAR-10/100 dataset, the local epoch is 10. For the Clothing1M dataset, the local epoch is set to 5. The learning rate is 0.01 for the CIFAR-10/100 dataset and 0.001 for the Clothing1M dataset. The batch size is 128 for CIFAR-10/100 and 32 for Clothing1M. Other settings remain the same for both datasets, including 10 rounds for pre-training, a total of 200 communication rounds, and an optimizer of SGD with a weight decay of 5e-4 and momentum of 0.9.