Adversarial Graph Embedding for Ensemble Clustering
Authors: Zhiqiang Tao, Hongfu Liu, Jun Li, Zhaowen Wang, Yun Fu
IJCAI 2019 | Conference PDF | Archive PDF | Plain Text | LLM Run Details
| Reproducibility Variable | Result | LLM Response |
|---|---|---|
| Research Type | Experimental | Experiments on eight real-world datasets are presented to show the effectiveness of AGAE over several state-of-the-art deep embedding and ensemble clustering methods. |
| Researcher Affiliation | Collaboration | Zhiqiang Tao1 , Hongfu Liu2 , Jun Li3 , Zhaowen Wang4 and Yun Fu1,5 1Department of Electrical and Computer Engineering, Northeastern University, Boston, MA 2Michtom School of Computer Science, Brandeis University, Waltham, MA 3Institute of Medical Engineering and Science, Massachusetts Institute of Technology, Cambridge, MA 4Adobe Research, Adobe Systems Incorporated, San Jose, CA 5Khoury College of Computer and Information Sciences, Northeastern University, Boston, MA |
| Pseudocode | Yes | Algorithm 1. Training of Adversarial Graph Auto-Encoder |
| Open Source Code | No | The paper does not contain an explicit statement about open-sourcing the code for the described methodology, nor does it provide a link to a code repository. |
| Open Datasets | Yes | 1https://archive.ics.uci.edu/ml/datasets; 2http://www.cad.zju.edu.cn/home/dengcai/; 3http://glaros.dtc.umn.edu/gkhome/; 4https://cs.stanford.edu/ acoates/stl10/; TDT2 [Cai et al., 2009]; Dslr [Kulis et al., 2011]; AWA4K [Lampert et al., 2009]; FRGC [Yang et al., 2016]. |
| Dataset Splits | No | The paper uses terms like 'pre-training' and 'fine-tuning' and mentions evaluation on datasets, but it does not specify explicit train/validation/test splits (e.g., percentages or sample counts) needed for reproduction. |
| Hardware Specification | No | The paper does not explicitly describe the specific hardware used for running its experiments (e.g., GPU models, CPU models, memory details). |
| Software Dependencies | No | The proposed AGAE was implemented with Tensor Flow toolbox. No specific version numbers for TensorFlow or other key software components are provided, which is necessary for reproducibility. |
| Experiment Setup | Yes | In Eq. (2), we set τ = 0.4 as default. In our model, we employed a two-layer GCN network as the probabilistic encoder (i.e., the generator), and a two-layer MLP network as the discriminator, where we set network dimensions of encoder as d-64-32 (i.e., m1=64 and m2=32) and discriminator as 32-128-1. ... We performed 200 epochs pre-training and fine-tuning on the USPS and STL-10 dataset, and 50 epochs for the remainder. The learning rate was set to be 1e-3 for pre-training and decayed to 1e-4 in fine-tuning. |