Table 1. Clustering quality measures (see text). TC is trajectory clustering, KM is K-means and JK is hierarchical jackknife.
| Number of interval | Number of manual Cluster | Other cluster | Number of other Cluster | Percent of other Cluster to manual Cluster | Discrimination index |
|---|---|---|---|---|---|
|
|
|||||
| 4 | 73 | TC | 72 | 55.1 | 0.80 |
| 4 | 20 | TC | 20 | 61.0 | 0.75 |
| 2 | 9 | TC | 9 | 77.4 | 0.56 |
| 4 | 20 | KM | 20 | 41.0 | 0.78 |
| 2 | 9 | KM | 9 | 28.6 | 0.59 |
| 4 | 20 | JK | 20 | 39.4 | 0.59 |
| 2 | 9 | JK | 9 | 30.0 | 0.34 |
| Using mean difference values | |||||
| 4 | 20 | KM | 20 | 34.6 | 0.47 |
| 2 | 9 | KM | 9 | 32.9 | 0.42 |
| 4 | 20 | JK | 20 | 26.9 | 0.57 |
| 2 | 9 | JK | 9 | 43.9 | 0.48 |