A Higher-Order Motif-Based Spatiotemporal Graph Imputation Approach for Transportation Networks

Table 3

Analysis of the imputation with MGIA (30 min time interval). The acronyms “r-missing” and “c-missing” represent the patterns of random data missing and continuous data missing, respectively.

Pattern	Metrics	Missing rate
		0.1	0.2	0.3	0.4	0.5	0.6
r-missing	RMSE	0.0548	0.0625	0.072	0.0939	0.121	0.1231
c-missing	RMSE	0.048	0.0534	0.0657	0.0854	0.0946	0.1106
r-missing	MAE	0.0492	0.0522	0.062	0.0821	0.0862	0.0974
c-missing	MAE	0.0412	0.051	0.052	0.0731	0.082	0.0967
r-missing	MAPE	3.71%	3.9%	4.47%	6.24%	6.37%	7.11%
c-missing	MAPE	3.24%	3.74%	3.96%	5.61%	6.23%	6.98%