Notice: The reproducibility variables underlying each score are classified using an automated LLM-based pipeline, validated against a manually labeled dataset. LLM-based classification introduces uncertainty; scores should be interpreted as estimates. Full accuracy metrics and methodology are described in [1]
Asynchronous Federated Clustering with Unknown Number of Clusters
Authors: Yunfan Zhang, Yiqun Zhang, Yang Lu, Mengke Li, Xi Chen, Yiu-ming Cheung
AAAI 2025 | Venue PDF | LLM Run Details
| Reproducibility Variable | Result | LLM Response |
|---|---|---|
| Research Type | Experimental | Extensive experiments demonstrate the efficacy of AFCL. The paper includes sections like 'Experiments', 'Experimental Setup', 'Visualization', 'Convergence Evaluation', and 'Clustering Performance Evaluation', comparing AFCL with other methods on 13 datasets and using metrics like SC and CH. |
| Researcher Affiliation | Academia | All authors are affiliated with universities: 'Guangdong University of Technology', 'Xiamen University', 'Shenzhen University', and 'Hong Kong Baptist University'. Email domains are primarily academic (.edu.cn, .xmu.edu.cn, .comp.hkbu.edu.hk), with one personal email (`@gmail.com`) associated with a university affiliation. |
| Pseudocode | Yes | The paper contains a clearly labeled algorithm block: 'Algorithm 1: AFCL: Asynchronous Federated Clustering.' |
| Open Source Code | Yes | The paper provides a direct link to a source-code repository: 'Code https://github.com/Yunfan-Zhang/AFCL' |
| Open Datasets | Yes | The paper states: '13 datasets, including two Gaussian spherical synthetic datasets, and 11 public real datasets collected from the UCI machine learning repository (Asuncion and Newman 2007), are utilized for the experiments, and their statistics are shown in Table 2.' |
| Dataset Splits | Yes | The paper describes how the data is partitioned for the federated setup: 'For each of the 20 trails, we first implement k-means with k = 5, to divide the whole dataset into five clients to simulate the extreme nonoverlapping distributions of non-IID clients. Then to simulate the asynchronous participation of clients, we randomly set each client with a different participation probability for controlling their upload in each iteration during the learning.' |
| Hardware Specification | No | The paper does not provide specific hardware details (e.g., GPU/CPU models, memory amounts) used for running its experiments. |
| Software Dependencies | No | The paper does not provide specific ancillary software details with version numbers. |
| Experiment Setup | Yes | The paper lists key hyperparameters in 'Algorithm 1: AFCL: Asynchronous Federated Clustering' as 'Input: X{1}, X{2}, ..., X{p}, k, ξ, η.' It further specifies for 'k': 'the initial number of seed points k is randomly selected from the range [k , 2k ].' It defines 'ξ' as 'a hyper-parameter controlling the sensitivity of balance weight w.r.t. the communication frequencies Θ.' and 'η' as 'the learning rate.' |