Dual Calibration-based Personalised Federated Learning

Authors: Xiaoli Tang, Han Yu, Run Tang, Chao Ren, Anran Li, Xiaoxiao Li

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

Reproducibility Variable Result LLM Response
Research Type Experimental Extensive experiments on widely used benchmark datasets demonstrate that DC-PFL outperforms eight state-of-the-art methods, surpassing the best-performing baseline by 1.22% and 9.22% in terms of accuracy on datasets CIFAR-10 and CIFAR-100, respectively.
Researcher Affiliation Academia 1College of Computing and Data Science, Nanyang Technological University, Singapore 2South China University of Technology, China 3Department of Electrical and Computer Engineering, The University of British Columbia, Canada
Pseudocode Yes Algorithm 1 DC-PFL
Open Source Code No The paper does not contain an explicit statement or link indicating that the source code for the described methodology is publicly available.
Open Datasets Yes We assess the performance of the proposed DC-PFL alongside baselines on datasets CIFAR-10 and CIFAR-100^1. ... 1https://www.cs.toronto.edu/ kriz/cifar.html
Dataset Splits Yes Furthermore, the data from each client is partitioned into three distinct subsets: training, evaluation, and testing, with an 8:1:1 allocation ratio.
Hardware Specification No The paper does not provide specific hardware details (e.g., GPU/CPU models, memory amounts, or detailed computer specifications) used for running its experiments.
Software Dependencies No The paper does not provide specific ancillary software details with version numbers (e.g., library or solver names with version numbers like Python 3.8, CPLEX 12.4) needed to replicate the experiment.
Experiment Setup Yes We optimize FL hyperparameters through an extensive grid search by adjusting the batch size for local training from {32, 64, 128, 256, 512} and the number of local training epochs from {1, 10, 30, 50, 100}. We utilize the SGD optimizer with a fixed learning rate (η) of 0.01 for both local training and global classifier training. The total number of communication rounds (T) is set to 100 on CIFAR-10 and to 500 on CIFAR100 to ensure convergence across all algorithms.