Matches in SemOpenAlex for { <https://semopenalex.org/work/W3119376013> ?p ?o ?g. }
Showing items 1 to 83 of
83
with 100 items per page.
- W3119376013 abstract "The problem of local minimum in k-means clustering, is commonly addressed by running the algorithm repeatedly in order to choose the best run. Although effective, the approach is computationally expensive. In this paper, we observe that the approach is effectively a comparison among different initializations. Thus, if there is a way to compare these initializations ab initio, there will be no need for repeated clustering. We propose such a technique in this paper. Specifically, we choose the initialization with the largest minimum inter-center distance (MIND), as the ‘best’ one. In other words, our technique is a general approach to improving existing seeding techniques. We demonstrate the concept with MIND-optimized versions of two standard algorithms: k-means and k-means++. Experiments show that in addition to drastic efficiency gain when compared to repetitive k-means, our approach improves the accuracy of the standard versions of these algorithms." @default.
- W3119376013 created "2021-01-18" @default.
- W3119376013 creator A5041759534 @default.
- W3119376013 creator A5066130528 @default.
- W3119376013 creator A5077115288 @default.
- W3119376013 date "2020-11-14" @default.
- W3119376013 modified "2023-10-12" @default.
- W3119376013 title "k- Means-MIND: An Efficient Alternative to Repetitive k-Means Runs" @default.
- W3119376013 cites W130123586 @default.
- W3119376013 cites W1590311885 @default.
- W3119376013 cites W1602564330 @default.
- W3119376013 cites W1987971958 @default.
- W3119376013 cites W1990772655 @default.
- W3119376013 cites W1998325344 @default.
- W3119376013 cites W2011430131 @default.
- W3119376013 cites W2024096326 @default.
- W3119376013 cites W2029698681 @default.
- W3119376013 cites W2071949631 @default.
- W3119376013 cites W2075098580 @default.
- W3119376013 cites W2122109176 @default.
- W3119376013 cites W2142458924 @default.
- W3119376013 cites W2150593711 @default.
- W3119376013 cites W2160558018 @default.
- W3119376013 cites W2606189485 @default.
- W3119376013 cites W2782722855 @default.
- W3119376013 cites W2783582361 @default.
- W3119376013 cites W2904777094 @default.
- W3119376013 cites W2907783748 @default.
- W3119376013 cites W2937352409 @default.
- W3119376013 cites W2940320374 @default.
- W3119376013 cites W2977992705 @default.
- W3119376013 cites W2990846584 @default.
- W3119376013 cites W2996740285 @default.
- W3119376013 cites W3008204019 @default.
- W3119376013 cites W3027970203 @default.
- W3119376013 cites W3082582247 @default.
- W3119376013 cites W3082718960 @default.
- W3119376013 cites W4235169531 @default.
- W3119376013 cites W944318169 @default.
- W3119376013 doi "https://doi.org/10.1109/iscmi51676.2020.9311598" @default.
- W3119376013 hasPublicationYear "2020" @default.
- W3119376013 type Work @default.
- W3119376013 sameAs 3119376013 @default.
- W3119376013 citedByCount "7" @default.
- W3119376013 countsByYear W31193760132022 @default.
- W3119376013 countsByYear W31193760132023 @default.
- W3119376013 crossrefType "proceedings-article" @default.
- W3119376013 hasAuthorship W3119376013A5041759534 @default.
- W3119376013 hasAuthorship W3119376013A5066130528 @default.
- W3119376013 hasAuthorship W3119376013A5077115288 @default.
- W3119376013 hasConcept C11413529 @default.
- W3119376013 hasConcept C114466953 @default.
- W3119376013 hasConcept C119857082 @default.
- W3119376013 hasConcept C154945302 @default.
- W3119376013 hasConcept C199360897 @default.
- W3119376013 hasConcept C207968372 @default.
- W3119376013 hasConcept C41008148 @default.
- W3119376013 hasConcept C73555534 @default.
- W3119376013 hasConceptScore W3119376013C11413529 @default.
- W3119376013 hasConceptScore W3119376013C114466953 @default.
- W3119376013 hasConceptScore W3119376013C119857082 @default.
- W3119376013 hasConceptScore W3119376013C154945302 @default.
- W3119376013 hasConceptScore W3119376013C199360897 @default.
- W3119376013 hasConceptScore W3119376013C207968372 @default.
- W3119376013 hasConceptScore W3119376013C41008148 @default.
- W3119376013 hasConceptScore W3119376013C73555534 @default.
- W3119376013 hasLocation W31193760131 @default.
- W3119376013 hasOpenAccess W3119376013 @default.
- W3119376013 hasPrimaryLocation W31193760131 @default.
- W3119376013 hasRelatedWork W12026041 @default.
- W3119376013 hasRelatedWork W13025791 @default.
- W3119376013 hasRelatedWork W13640671 @default.
- W3119376013 hasRelatedWork W1813111 @default.
- W3119376013 hasRelatedWork W3970919 @default.
- W3119376013 hasRelatedWork W5223388 @default.
- W3119376013 hasRelatedWork W619728 @default.
- W3119376013 hasRelatedWork W7081844 @default.
- W3119376013 hasRelatedWork W8413239 @default.
- W3119376013 hasRelatedWork W8941019 @default.
- W3119376013 isParatext "false" @default.
- W3119376013 isRetracted "false" @default.
- W3119376013 magId "3119376013" @default.
- W3119376013 workType "article" @default.