Robust Multi-View Spectral Clustering via Low-Rank and Sparse Decomposition

Authors: Rongkai Xia, Yan Pan, Lei Du, Jian Yin

AAAI 2014 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental Experimental results on various real world datasets show that the proposed method has superior performance over several state-of-the-art methods for multi-view clustering.
Researcher Affiliation Academia Rongkai Xia, Yan Pan, Lei Du, and Jian Yin Sun Yat-sen University, Guangzhou, China
Pseudocode Yes Algorithm 1 Algorithm for transition matrix construction
Open Source Code No The paper does not provide any statement about releasing open-source code or a link to a code repository for the described methodology.
Open Datasets Yes We report the experimental results on six real-world datasets: BBC and BBCSport1 for news article clustering, Web KB (Sindhwani, Niyogi, and Belkin 2005) for webpages clustering, UCI digits (Asuncion and Newman 2007) and Flower172 for image clustering, and Columbia Consumer Video (CCV) (Jiang et al. 2011) for video event clustering.
Dataset Splits No The paper does not explicitly provide training, validation, and test dataset splits with specific percentages, counts, or references to predefined splits.
Hardware Specification No The paper does not explicitly describe the specific hardware (e.g., GPU/CPU models, memory) used to run its experiments.
Software Dependencies No The paper does not provide specific version numbers for any software dependencies. It mentions 'Gaussian kernels' and 'k-means clustering' but without specific library or version information.
Experiment Setup Yes In all the experiments, Gaussian kernels are used to build the similarity matrix for each single view. The standard deviation is set to the median of the pairwise Euclidean distances between every pair of data points for all of the datasets except BBC and BBCSport. For the BBC and BBCSport datasets, we follow (Kumar and Daum e 2011) to set the standard deviation to be 100. For MMC (Zhou and Burges 2007) and the proposed RMSC, the transition probability matrix for each view is constructed by P = D 1S, where S is the similarity matrix and D is a diagonal matrix with Dii being the sum of the elements of the ith row in S. In RMSC, the regularization parameter λ is set to be 0.005 3.