Localized Incomplete Multiple Kernel k-means

Authors: Xinzhong Zhu, Xinwang Liu, Miaomiao Li, En Zhu, Li Liu, Zhiping Cai, Jianping Yin, Wen Gao

IJCAI 2018 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental Comprehensive experiments demonstrate that our algorithm significantly outperforms the state-of-the-art comparable algorithms proposed in the recent literature, verifying the advantage of considering local structure. and 4 Experiments section
Researcher Affiliation Academia Xinzhong Zhu1,7, Xinwang Liu2, Miaomiao Li2, En Zhu2, Li Liu3,4, Zhiping Cai2, Jianping Yin5 and Wen Gao6 1 College of Mathematics, Physics and Information Engineering, Zhejiang Normal University, China 2 School of Computer, National University of Defense Technology, Changsha, China 3 College of System Engineering, National University of Defense Technology, Changsha, China 4 University of Oulu, Finland 5 Dongguan University of Technology, Guangdong, China 6 School of Electronics Engineering and Computer Science, Peking University, Beijing, China 7 School of Electronic Engineering, XIDIAN University, Xi an, Shanxi, China
Pseudocode Yes Algorithm 1 Proposed LI-MKKM
Open Source Code No The paper does not provide any statement about releasing source code, nor does it include a link to a code repository.
Open Datasets Yes The proposed algorithm is experimentally evaluated on eight widely used MKL benchmark data sets shown in Table 2. They are Oxford Flower17 and Flower1021 and Caltech1022. For these datasets, all kernel matrices are pre-computed and can be publicly downloaded from the above websites. 1http://www.robots.ox.ac.uk/ vgg/data/flowers/ 2http://files.is.tue.mpg.de/pgehler/projects/iccv09/
Dataset Splits No The paper describes generating missing patterns and repeating experiments with random initialization, and that the true number of clusters is known, but it does not specify explicit training, validation, or test dataset splits or percentages for data partitioning.
Hardware Specification No The paper discusses computational complexity and running time but does not provide any specific details about the hardware (e.g., GPU/CPU models, memory) used for the experiments.
Software Dependencies No The paper does not specify any software dependencies with version numbers (e.g., programming languages, libraries, or frameworks with their specific versions) used for the experiments.
Experiment Setup Yes For all data sets, it is assumed that the true number of clusters is known and it is set as the true number of classes. We follow the approach in [Liu et al., 2017a] to generate the missing vectors {sp}m p=1. The parameter ε, termed missing ratio in this experiment, controls the percentage of samples that have absent views, and it affects the performance of the algorithms in comparison. Specifically, ε on all the four data sets is set as [0.1 : 0.1 : 0.9]. and In all the above experiments, we empirically set τ = 0.1 for all datasets