The Complexity of k-Means Clustering when Little is Known

Authors: Robert Ganian, Thekla Hamm, Viktoriia Korchemna, Karolina Okrasa, Kirill Simonov

ICML 2022 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Theoretical Here, we study the complexity of k-means clustering in settings where most of the data is not known or simply irrelevant. To obtain a more fine-grained understanding of the tractability of this clustering problem, we apply the parameterized complexity paradigm and obtain three new algorithms for k-means clustering of incomplete data...
Researcher Affiliation Academia 1Algorithms and Complexity Group, TU Wien, Austria 2Institute of Informatics, University of Warsaw, Poland.
Pseudocode No The paper describes algorithmic procedures in prose and through mathematical notation within proofs, but it does not contain structured pseudocode or algorithm blocks (e.g., a clearly labeled Algorithm 1).
Open Source Code No The paper does not provide concrete access to source code for the methodology described.
Open Datasets No The paper is theoretical and does not conduct experiments with a dataset. While it mentions the Netflix Prize dataset as a motivating example ('The dataset is available at https://www.kaggle.com/netflix-inc/netflix-prize-data.'), it is not 'used in the experiments' as no experiments are reported.
Dataset Splits No The paper focuses on theoretical complexity and algorithm design, and thus does not describe experimental validation or dataset splits.
Hardware Specification No The paper is theoretical and does not describe experimental procedures, therefore no specific hardware specifications are mentioned.
Software Dependencies No The paper is theoretical and focuses on algorithm design and complexity, and thus does not specify any software dependencies with version numbers.
Experiment Setup No The paper focuses on theoretical aspects of k-means clustering and does not describe any specific experimental setup details or hyperparameters.