New l2,1-Norm Relaxation of Multi-Way Graph Cut for Clustering
Authors: Xu Yang, Cheng Deng, Xianglong Liu, Feiping Nie
AAAI 2018 | Conference PDF | Archive PDF | Plain Text | LLM Run Details
| Reproducibility Variable | Result | LLM Response |
|---|---|---|
| Research Type | Experimental | Experimental results on several benchmark datasets show that our method significantly outperforms several state-of-the-art clustering approaches. |
| Researcher Affiliation | Academia | 1School of Electronic Engineering, Xidian University, Xian 710071, China 2Beihang University, Beijing 100191, China 3Northwestern Polytechnical University, Xian 710072, China |
| Pseudocode | Yes | Algorithm 1 Algorithm to solve the problem (25) [...] Algorithm 2 Algorithm to solve the problem (12) |
| Open Source Code | No | The paper does not provide any statement about making its source code available or a link to a code repository. |
| Open Datasets | Yes | Two UCI datasets are Dermatol and Yeast (Asuncion and Newman 2007). Two object datasets COIL20 and COIL100 (Nene et al. 1996) contain different objects imaged at every angle in a 360 rotation and the backgrounds have been discarded. [...] Yale contains 165 gray-scale images in 15 individuals, one per different facial expression or configuration. Yale B (Georghiades, Belhumeur, and Kriegman 2001) has 38 individuals and around 64 near frontal images under different illuminations per individual. [...] Umist (Graham and Allinson 1998) consists of 20 individuals [...]. MSRA (Liu et al. 2007) contains 12 individuals with different background and illumination conditions. |
| Dataset Splits | No | The paper does not specify explicit training, validation, or test dataset splits using percentages, sample counts, or references to predefined splits for reproducibility. It mentions 'synthetic data' and 'nine benchmark datasets' but does not detail how these were partitioned for training/validation/testing. |
| Hardware Specification | No | The paper does not specify any hardware details such as GPU/CPU models, processors, or memory used for running the experiments. |
| Software Dependencies | No | The paper does not provide specific software dependencies or their version numbers (e.g., Python, PyTorch, TensorFlow versions, or specific solver versions). |
| Experiment Setup | Yes | We construct the original weight matrix W with probabilistic K-nearest neighbors for each datasets. The weight Wij is calculated as nearest-neighbor graph (Gu and Zhou 2009), and the number of neighbors is set to 5. We adopt K-means to refine the final results, and repeat K-means for 50 times with random initializations. |