Skip to main content

Advances, Systems and Applications

Journal of Cloud Computing Cover Image

Table 2 Comparison of running time

From: Efficient parallel spectral clustering algorithm design for large data sets under cloud computing environment

Data volume Machines Similar matrix (sec) Eigenvector (sec) K-means (sec) Total time (sec)
DS1 (10000) 1 0.386 0.481 0.156 1.023
2 0.532 1.099 3.436 5.067
4 0.186 0.364 1.137 1.687
6 0.096 0.175 0.582 0.853
8 0.038 0.065 0.231 0.334
10 0.025 0.050 0.204 0.279
DS2 (50000) 1 7.251 9.315 2.947 19.513
2 9.814 12.879 4.139 26.832
4 3.162 3.963 1.376 8.501
6 2.299 2.829 1.239 6.367
8 1.477 1.881 0.910 4.268
10 1.218 1.555 0.887 3.660
DS3 (100000) 1 19.228 23.982 8.572 51.782
2 11.234 14.409 4.414 30.057
4 5.736 6.538 2.246 14.520
6 4.007 5.587 1.432 11.026
8 2.965 4.056 0.901 7.922
10 2.359 3.453 0.654 6.466
DS4 (1000000) 1 7671.580 9603.573 3422.564 20697.717
2 37.590 46.058 16.678 100.326
4 19.629 23.755 8.719 53.103
6 10.126 18.473 6.475 35.074
8 8.797 13.532 4.865 27.194
10 6.894 11.415 3.852 22.161
DS5 (5000000) 1 31602.604 39909.984 16630.820 88143.408
2 150.853 191.559 80.850 423.262
4 75.164 98.906 39.213 213.283
6 50.273 70.427 22.087 142.787
8 40.032 53.142 18.521 111.695
  10 30.841 42.112 13.940 86.893