<abstract xmlns="http://www.w3.org/1999/xhtml"><p>Clustering as a fundamental unsupervised learning is considered an important method of data analysis, and <italic>K</italic>-means is demonstrably the most popular clustering algorithm. In this paper, we consider clustering on feature space to solve the low efficiency caused in the Big Data clustering by <italic>K</italic>-means. Different from the traditional methods, the algorithm guaranteed the consistency of the clustering accuracy before and after descending dimension, accelerated <italic>K</italic>-means when the clustering centeres and distance functions satisfy certain conditions, completely matched in the preprocessing step and clustering step, and improved the efficiency and accuracy. Experimental results have demonstrated the effectiveness of the proposed algorithm.</p></abstract>

Clustering as a fundamental unsupervised learning is considered an important method of data analysis, and K-means is demonstrably the most popular clustering algorithm. In this paper, we consider clustering on feature space to solve the low efficiency caused in the Big Data clustering by K-means. Different from the traditional methods, the algorithm guaranteed the consistency of the clustering accuracy before and after descending dimension, accelerated K-means when the clustering centeres and distance functions satisfy certain conditions, completely matched in the preprocessing step and clustering step, and improved the efficiency and accuracy. Experimental results have demonstrated the effectiveness of the proposed algorithm.

Improvement of the Fast Clustering Algorithm Improved by <italic>K</italic>-Means in the Big Data

Applied Mathematics and Nonlinear Sciences

This work is licensed under the Creative Commons Attribution 4.0 International License.

Improvement of the Fast Clustering Algorithm Improved by -Means in the Big Data

Clustering as a fundamental unsupervised learning is considered an important method of data analysis, and K-means is demonstrably the most popular clustering...

Name	K-means	Spherical K-means	K-medoids
ORL	10.77	8.45	2.74
YALE	8.88	2.89	3.96
COIL20	8.86	6.22	5.70
CMD	10.22	4.86	6.61
DLBCL	7.70	4.76	6.92
LunG	5.13	1.21	1.96
Prostate	15.47	10.96	15.44

Name	d	n	k
ORL	4,096	400	20
YALE	4,096	165	15
COIL20	16,384	1,440	20
CMD	7,129	60	2
DLBCL	7,129	77	2
LunG	1,000	197	4
Prostate	12,600	102	2

Name	K-means	Spherical K-means	K-medoids
ORL	2.309e–14	1.705e–13	1.243e–13
YALE	4.263e–14	2.398e–14	7.105e–15
COIL20	2.757e–12	1.121e–11	4.547e–13
CMD	7.105e–15	6.128e–14	1.776e–14
DLBCL	7.105e–15	9.548e–14	3.730e–14
LunG	3.908e–14	4.796e–14	3.553e–14
Prostate	7.105e–15	1.172e–13	3.553e–14

Size	K-means	Spherical K-means	K-medoids
d= 1,000	0	0	0
d= 2,000	1.863e–09	0	0
d= 5,000	3.725e–09	2.842e–13	3.275e–09
d= 10,000	3.725e–09	1.137e–13	7.451e–09
d= 20,000	1.490e–08	6.253e–13	0
d= 50,000	2.980e–08	0	5.960e–08

Size	K-means	Spherical K-means	K-medoids
d= 1,000	1.01	1.05	1.00
d= 2,000	1.46	1.13	1.30
d= 5,000	4.30	2.59	2.73
d= 10,000	7.97	4.71	5.19
d= 20,000	17.76	8.42	10.09
d= 50,000	44.10	33.60	26.30

Improvement of the Fast Clustering Algorithm Improved by K-Means in the Big Data

Data publikacji: 20 sty 2020

Zakres stron: 1 - 10

Otrzymano: 23 wrz 2019

Przyjęty: 26 gru 2019

DOI: https://doi.org/10.2478/amns.2020.1.00001

Słowa kluczowe
Big Data, Clustering, -means, Feature space

© 2020 Ting Xie et al., published by Sciendo

This work is licensed under the Creative Commons Attribution 4.0 International License.

Figure 1

Figure 2

Figure 3

The comparison of the run times on the actual data

The date information in the algorithm test

The comparison of the objective functions on the actual data

The comparison of the objective functions on the artificial data

The comparison of the run time on the artificial data

Improvement of the Fast Clustering Algorithm Improved by K-Means in the Big Data

Data publikacji: 20 sty 2020

Zakres stron: 1 - 10

Otrzymano: 23 wrz 2019

Przyjęty: 26 gru 2019

DOI: https://doi.org/10.2478/amns.2020.1.00001

Słowa kluczoweBig Data, Clustering, -means, Feature space

© 2020 Ting Xie et al., published by Sciendo

This work is licensed under the Creative Commons Attribution 4.0 International License.

Figure 1

Figure 2

Figure 3

The comparison of the run times on the actual data

The date information in the algorithm test

The comparison of the objective functions on the actual data

The comparison of the objective functions on the artificial data

The comparison of the run time on the artificial data

Słowa kluczowe
Big Data, Clustering, -means, Feature space