Table 4 MSD comparisons with different number of points when k = 5 (in millions of actual values).
From: A practical comparison of two K-Means clustering algorithms
for k= 5 | 50 pts | 100 pts | 1000 pts | 5000 pts | 10000 pts |
|---|---|---|---|---|---|
Lloyd min MSD | 4.078 | 5.222 | 8.116 | 7.997 | 8.089 |
Lloyd global avg. MSD | 7.142 | 7.423 | 9.604 | 9.915 | 9.804 |
Lloyd max MSD | 9.622 | 9.250 | 11.439 | 11.259 | 10.875 |
Progressive min MSD | 7.070 | 5.750 | 7.539 | 8.031 | 8.071 |
Progressive global avg. MSD | 8.610 | 8.395 | 9.715 | 9.916 | 9.804 |
Progressive max MSD | 10.442 | 10.240 | 11.761 | 11.247 | 10.859 |