Solving Soft Clustering Ensemble via $k$-Sparse Discrete Wasserstein Barycenter

Authors: Ruizhe Qin, Mengying Li, Hu Ding

NeurIPS 2021 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental Finally, we conduct the experiments to evaluate our proposed algorithms.
Researcher Affiliation Academia 1School of Computer Science and Technology 2School of Data Science University of Science and Technology of China red46@mail.ustc.edu.cn, limengy@mail.ustc.edu.cn, huding@ustc.edu.cn
Pseudocode Yes Algorithm 1 (1 + ϵ)-Approximate SCE Algorithm
Open Source Code No The paper does not provide an explicit statement or link to open-source code for the described methodology.
Open Datasets Yes We consider three real datasets: USPS has 11000 data items in R256 with k = 10 [32]; IRIS [22] has 150 data items in R4 with k = 3; CIFAR-10 [38] has 10000 data items in R3072 with k = 10.
Dataset Splits No The paper describes the datasets used and how clustering solutions were generated (random projections, k-means) but does not specify training, validation, or test splits for the datasets themselves.
Hardware Specification Yes All the experimental results were obtained on a server equipped with 2.8GHz Intel CPU, 8GB main memory, and Matlab 2019a.
Software Dependencies Yes All the experimental results were obtained on a server equipped with 2.8GHz Intel CPU, 8GB main memory, and Matlab 2019a.
Experiment Setup Yes We set m = 1000 (i.e., the number of generated clustering solutions for ensemble)... Similar with [24, 13], we apply random projections to generate the clustering solutions (in each random subspace, we use k-means to cluster the data). ... Our sampling idea of Section 4.2 is incorporated into the alternating minimization Wasserstein barycenter algorithm [62], which is denoted as AM-r with r representing the sample rate... All the results are averaged across 30 trials.