Multi-View Clustering and Semi-Supervised Classification with Adaptive Neighbours

Authors: Feiping Nie, Guohao Cai, Xuelong Li

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

Reproducibility Variable Result LLM Response
Research Type Experimental Extensive experimental results on different real-world datasets show that the proposed model outperforms other state-of-the-art multi-view algorithms.
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
Pseudocode Yes Algorithm 1 Multi-view Learning with Adaptive Neighbours
Open Source Code No The paper does not provide any specific links or explicit statements regarding the public availability of the source code for the described methodology.
Open Datasets Yes MSRC-v1 data set contain 240 images... Following (Cai et al. 2011), we select 7 classes... Handwritten numerals (HW) data set is comprised of 2,000 data points... Caltech101 is an object recognition data set... We follow previous work (Li et al. 2015)... NUS-WIDE is a real-world web image dataset...
Dataset Splits No The paper states, 'In terms of semi-supervised classification, we choose the front 20% data for as labeled sample to mimic the real situation (l u).' However, it does not provide specific details on a separate validation split for hyperparameter tuning or model selection.
Hardware Specification No The paper does not provide any specific details about the hardware used for running the experiments.
Software Dependencies No The paper does not specify any software dependencies with version numbers.
Experiment Setup Yes There is only one parameter λ in our model... initialize λ to a random positive value, 1 to 30 in our experiment, and decrease it (λ = λ/4) if the connected components of S is greater than class number c or increase it (λ = λ 4) if smaller than c in each iteration... To all dataset, each sample is assigned 9 nearest neighbours to construct graph. In terms of semi-supervised classification, we choose the front 20% data for as labeled sample to mimic the real situation (l u)... we perform 50 times experiments for all methods on each dataset.