Figures & data
Figure 1. Study area (Shanghai, China) (a), and spatial distribution of check-in data by activities (b).
Figure 2. Flow chart of the proposed method.
Figure 3. Region-TTD matrix.
Figure 4. Distribution of singular values.
Figure 5. Reconstruction accuracy compared with the observed matrix.
Figure 6. Reconstruction accuracy compared to original temporal variation of travel demands.
Figure 7. TTD characteristics of the top six LSTAS.
Figure 8. Spatial distributions of the top six LSTAS.
Figure 9. Cluster validity indices of the K-means algorithm via (a) Dunn’s and Silhouette index, and (b) Davies-Bouldin index.
Figure 10. Similarities among each TTD and centers of the 5 Clusters. (a) Cluster #1; (b) Cluster #2; (c) Cluster #3; (d) Cluster #4. (e) Cluster #5.
Figure 11. Result of region aggregation and concentric urban form revealed by 5 Clusters. (a) six circles with the radii from 1 to17 km; (b) proportions of various clusters in zones.
