Data-Driven Approach for Passenger Mobility Pattern Recognition Using Spatiotemporal Embedding

Table 3

The number of clusters and algorithm performance based on different parameters.

ID	δ	ε	K	SSE	SC
1	8	7	27	33136	0.466
2	8	7.5	29	33407	0.473
3	8	8	24	30895	0.439
4	8	8.5	18	34989	0.139
5	8	9	16	31018	0.621
6	8	9.5	15	31731	0.568
7	8	10	11	23133	0.712
8	10	7	24	28931	0.248
9	10	7.5	22	29168	0.38
10	10	8	19	27447	0.421
11	10	8.5	14	27925	0.361
12	10	9	13	27374	0.616
13	10	9.5	12	28828	0.621
14	10	10	9	27158	0.657
15	12	7	19	26676	0.295
16	12	7.5	17	26224	0.379
17	12	8	16	26786	0.424
18	12	8.5	11	26077	0.188
19	12	9	11	26775	0.614
20	12	9.5	10	26596	0.56
21	12	10	6	26638	0.686
22	14	7	18	25702	0.356
23	14	7.5	16	24084	0.381
24	14	8	12	25114	0.503
25	14	8.5	10	23429	0.597
26	14	9	9	23796	0.646
27	14	9.5	9	23889	0.596
28	14	10	3	23647	0.793
29	16	7	18	24333	0.348
30	16	7.5	14	22281	0.38
31	16	8	10	23117	0.502
32	16	8.5	10	23255	0.487
33	16	9	9	23760	0.646
34	16	9.5	6	23715	0.815
35	16	10	5	23568	0.596
36	18	7	17	22439	0.151
37	18	7.5	13	21892	0.364
38	18	8	10	23129	0.25
39	18	8.5	7	22538	0.507
40	18	9	7	23407	0.681
41	18	9.5	7	23436	0.629
42	18	10	6	23675	0.654