Research Article
The Impact of Truck Proportion on Traffic Safety Using Surrogate Safety Measures in China
Table 5
k-means clustering analysis in different periods.
| Different periods (min) | Clustering & testing | Classification | 85%V–15%V | Variation coefficient |
| 5 | Initial clustering center | 1 (dangerous) | 68.957 | 0.394 | | 2 (safe) | 11.923 | 0.090 | Final cluster center | 1 (dangerous) | 50.715 | 0.276 | | 2 (safe) | 32.290 | 0.172 | The value of split point | 41.503 | 0.224 | ANOVA | F | 1525.788 | 2388.247 | Sig. | 0.000 | 0.000 | |
| 10 | Initial clustering center | 1 (dangerous) | 64.489 | 0.348 | | 2 (safe) | 12.457 | 0.095 | Final clustering center | 1 (dangerous) | 51.001 | 0.273 | | 2 (safe) | 33.252 | 0.175 | The value of split point | 42.127 | 0.224 | ANOVA | F | 945.529 | 1733.678 | Sig. | 0.000 | 0.000 | |
| 15 | Initial clustering center | 1 (dangerous) | 64.067 | 0.371 | | 2 (safe) | 14.971 | 0.136 | Final clustering center | 1 (dangerous) | 51.266 | 0.271 | | 2 (safe) | 34.077 | 0.174 | The value of split point | 42.672 | 0.223 | ANOVA | F | 626.887 | 1129.481 | Sig. | 0.000 | 0.000 |
|
|