النتائج (
العربية) 2:
[نسخ]نسخ!
This section is devoted to studying how effective the proposed mobility prediction<br>approach is. We first mine the trajectory dataset DS500 which is produced by a<br>dataset generator into discover all mobility patterns which have support values higher than the predefined minimum support threshold. Second, all mobility patterns are<br>clustered into k groups and then generating mobility rules for each group of mobility<br>patterns. For evaluating, we use a test set TS10 which consists of 7 trajectories with<br>label 1, 2 trajectories with label 2 and 1 trajectories with label 3 from the dataset<br>DS500. The next location of each trajectory in T10 is predicted and the results are<br>used to calculate precision and recall. Varying k from 2 to 10 on 1 incremental steps,<br>we obtain experimental results as in Figure 3.
يجري ترجمتها، يرجى الانتظار ..
