Spectral Clustering with Brainstorming Process for Multi-View Data

Authors: Jeong-Woo Son, Junkey Jeon, Alex Lee, Sun-Joong Kim

AAAI 2017 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental Experimental results in three tasks show the effectiveness of the proposed method comparing with ordinary single and multi-view spectral clusterings.
Researcher Affiliation Academia Jeong-Woo Son, Junkey Jeon, Alex Lee, Sun-Joong Kim {jwson, jkjeon, lhjalex, kimsj}@etri.re.kr Smart Platform Research Department, Electronics and Telecommunications Research Institute 218 Gajeong-ro, Yuseong-gu, Deajeon, Korea
Pseudocode No The paper describes the proposed method and its procedures using prose and mathematical equations but does not include structured pseudocode or clearly labeled algorithm blocks.
Open Source Code No The paper does not provide any specific links or statements regarding the availability of source code for the described methodology.
Open Datasets Yes The evaluation of the proposed method is performed with three datasets: synthetic, Reuter multilingual, and BBC datasets. Reuter multilingual dataset is composed of documents from six categories (Amini, Usunier, and Goutte 2009)... BBC dataset contains 2,225 news articles in five categories (Greene and Cunningham 2006).
Dataset Splits No The paper describes how instances are sampled for the synthetic and Reuter multilingual datasets, but it does not specify any training, validation, or test dataset splits for any of the datasets used in the experiments.
Hardware Specification No The paper does not provide any specific hardware details such as GPU/CPU models, processor types, or memory amounts used for running the experiments.
Software Dependencies No The paper mentions techniques like k-means clustering, Gaussian kernel, and Latent Semantic Analysis, and refers to 'ordinary quadratic programmings', but it does not provide specific software dependencies with version numbers needed for replication.
Experiment Setup Yes In Gaussian kernel, we used the locality preserved kernel (Zelnik-Manor and Perona 2004) for the synthetic and Reuter multilingual datasets, while we give the standard deviation 100 for BBC dataset. ... MSE has a user parameter the size of neighbor nodes. ... we set it as 9 for the synthetic dataset and 5 for both Reuter multilingual and BBC datasets. ... In all experiments, the user parameter C in both discussion and conclusion steps is fixed as 1.0. Experimental results are obtained by 10 trials for all datasets.