Balanced Clustering via Exclusive Lasso: A Pragmatic Approach
Authors: Zhihui Li, Feiping Nie, Xiaojun Chang, Zhigang Ma, Yi Yang
AAAI 2018 | Conference PDF | Archive PDF | Plain Text | LLM Run Details
| Reproducibility Variable | Result | LLM Response |
|---|---|---|
| Research Type | Experimental | Extensive experiments on several large-scale datasets validate the advantage of the proposed algorithms compared to the state-of-the-art clustering algorithms. In this section, extensive experiments are conducted to evaluate the proposed clustering methods. |
| Researcher Affiliation | Collaboration | Zhihui Li,1 Feiping Nie,2 Xiaojun Chang,3 Zhigang Ma,3 Yi Yang4 1Beijing Etrol Technologies Co., Ltd. 2Centre for OPTical Imagery Analysis and Learning, Northwestern Polytechnical University. 3School of Computer Science, Carnegie Mellon University. 4Centre for Artificial Intelligence, University of Technology Sydney. |
| Pseudocode | Yes | Algorithm 1 Algorithm to solve the objective function of balanced k-means. Algorithm 2 Algorithm to solve the objective function of balanced min-cut. |
| Open Source Code | No | The paper does not provide any links to open-source code or state that code will be made available. |
| Open Datasets | Yes | MNIST Handwritten Digit Dataset: The MNIST handwritten digit dataset (Le Cun et al. 2011). Yale B face dataset: The Yale B dataset (Georghiades, Belhumeur, and Kriegman 2001). ORL face dataset: The ORL dataset (Samaria and Harter 1994). JAFFE Japanese Female Facial Expression dataset: The JAFFE dataset (Lyons, Budynek, and Akamatsu 1999). Human EVA Motion dataset 1. (1http://vision.cs.brown.edu/humaneva/). Coil20 Object dataset: We use the Coil20 dataset (Nene, Nayar, and Murase 1996). |
| Dataset Splits | No | The paper mentions that for MNIST, 'The dataset contains 60,000 training images and 10,000 testing images. We merge all the training and testing images in the experiments.' This indicates no separate validation split, and the existing test split was merged, so specific train/validation/test splits for reproducibility are not provided. |
| Hardware Specification | No | The paper does not specify any hardware details such as CPU, GPU models, or memory used for experiments. |
| Software Dependencies | No | The paper does not list any specific software dependencies with version numbers. |
| Experiment Setup | Yes | For the regularization parameter γ in Eq. (13) and Eq. (24), we tune them by a grid-search strategy from {10 6, 10 4, 10 2, 100, 102, 104, 106}. We similarly tune the regularization parameters of all the comparison algorithms from the aforementioned range. |