Lifelong Spectral Clustering
Authors: Gan Sun, Yang Cong, Qianqian Wang, Jun Li, Yun Fu5867-5874
AAAI 2020 | Conference PDF | Archive PDF | Plain Text | LLM Run Details
| Reproducibility Variable | Result | LLM Response |
|---|---|---|
| Research Type | Experimental | Finally, the empirical experiments on several real-world benchmark datasets demonstrate that our L2SC model can effectively improve the clustering performance when comparing with other state-of-the-art spectral clustering algorithms. |
| Researcher Affiliation | Academia | Gan Sun,1,3 Yang Cong,2 Qianqian Wang,3 Jun Li,4 Yun Fu3 1University of Chinese Academy of Sciences, China. 2State Key Laboratory of Robotics, Shenyang Institute of Automation, Chinese Academy of Sciences, China. 3Northeastern University, USA. 4MIT, USA. |
| Pseudocode | Yes | Algorithm 1 Lifelong Spectral Clustering (L2SC) Model |
| Open Source Code | No | The paper does not contain any statements about making source code publicly available or links to a code repository. |
| Open Datasets | Yes | We thus use two datasets in this paper: Web KB41 with 2500 dimensions and Reuters2 with 6370 dimensions, respectively. ... For the Cluster-inconsistent dataset, we also adopt 20News Groups3 dataset with 3000 dimensions by following (Zhang et al. 2018)... The footnotes provide the URLs: 1http://www.cs.cmu.edu/afs/cs.cmu.edu/project/theo20/www /data/, 2http://www.cad.zju.edu.cn/home/dengcai/Data/Text Data.html, 3http://qwone.com/ jason/20Newsgroups/. |
| Dataset Splits | No | The paper mentions "10 random repetitions" for averaging results and describes how categories are selected for tasks in the 20News Groups dataset (e.g., "1-th and 3-th tasks are set as 3 categories, and the 2-th and 4-th tasks are set as 4 categories"), but it does not provide specific train/validation/test dataset split percentages or absolute sample counts for any of the datasets. |
| Hardware Specification | Yes | We perform all the experiments on the computer with Intel i7 CPU, 8G RAM. |
| Software Dependencies | No | The paper states "We implement all the models in MATLAB," but it does not provide specific version numbers for MATLAB or any other software dependencies, libraries, or solvers. |
| Experiment Setup | Yes | all the used parameters of the models are tuned in {10^3 i}^10_{i=1} {10^2 i}^10_{i=2} {10^1 i}^10_{i=2} {2 i}^10_{i=1} {40 i}^20_{i=1}. and appropriate parameters can make the generalization performance better, e.g., λ = 100 for Web KB4 dataset. |