Unsupervised Large Graph Embedding
Authors: Feiping Nie, Wei Zhu, Xuelong Li
AAAI 2017 | Conference PDF | Archive PDF | Plain Text | LLM Run Details
| Reproducibility Variable | Result | LLM Response |
|---|---|---|
| Research Type | Experimental | Extensive experiments on several public available data sets demonstrate the efficiency and effectiveness of the proposed method. |
| Researcher Affiliation | Academia | 1School of Computer Science and Center for OPTical IMagery Analysis and Learning (OPTIMAL), Northwestern Polytechnical University, Xi an 710072, Shaanxi, P. R. China 2Center for OPTical IMagery Analysis and Learning (OPTIMAL), Xi an Institute of Optics and Precision Mechanics, Chinese Academy of Sciences, Xi an 710119, Shaanxi, P. R. China {feipingnie@gmail.com, zwvews@gmail.com, xuelong_li@opt.ac.cn} |
| Pseudocode | Yes | Algorithm 1 Large Graph Embedding Dimensionality Reduction |
| Open Source Code | No | The paper states 'All the codes in the experiments are implemented in MATLAB R2015b', but does not provide a link to the source code or an explicit statement about its public availability. |
| Open Datasets | Yes | We conduct experiments on 5 different public available data sets downloaded from the Lib SVM data sets page1, UCI machine learning repository 2, and Deng Cai s page 3. ... 1https://www.csie.ntu.edu.tw/~cjlin/libsvmtools/datasets/ 2http://archive.ics.uci.edu/ml/ 3http://www.cad.zju.edu.cn/home/dengcai/Data/data.html |
| Dataset Splits | No | The paper uses k-means for evaluation after dimensionality reduction but does not specify explicit training, validation, or testing splits for the datasets. |
| Hardware Specification | Yes | All the codes in the experiments are implemented in MATLAB R2015b , and run on a Windows 10 machine with 3.20 GHz i5-3470 CPU, 16 GB main memory. |
| Software Dependencies | Yes | All the codes in the experiments are implemented in MATLAB R2015b |
| Experiment Setup | Yes | The regularization parameter α is default set as 0.01 in both SR and ULGE. For parameter m, i.e. the number of anchors, used in ULGE, we empirically set m = 1000. To speed up ULGE-K, we suggest to perform down sampling to generate anchors, and in this paper, the decimation factor is set as 10 for all data sets except USPS which is set as 3. We use same Gaussian kernel for all LE, LPP and SR, and the reduced dimension is set as number of class of the data set. We use 5-nearest neighbor to construct graph for all the methods. |