Table 3 Comparison of different models on the highway traffic flow dataset.

From: Spatio-temporal transformer and graph convolutional networks based traffic flow prediction

Methods

PEMS04

PEMS07

PEMS08

MAE

RMSE

MAPE

MAE

RMSE

MAPE

MAE

RMSE

MAPE

LSTM

26.91

42.90

17.60

30.24

48.67

12.56

21.49

33.87

13.09

DCRNN

23.70

36.70

15.54

24.94

37.71

10.89

18.35

27.87

11.97

GWN

20.11

31.79

14.14

22.24

35.10

9.70

14.89

23.64

9.62

AGCRN

19.76

32.61

12.95

20.92

34.73

8.82

16.31

25.88

10.29

ASTGCN

21.25

33.65

13.93

24.53

38.12

10.55

18.27

28.43

11.05

STGCN

24.58

38.28

16.79

29.34

45.74

13.42

19.52

29.91

12.87

ASTGNN

18.32

31.00

12.29

19.23

32.79

8.43

12.96

22.83

8,85

STWave

18.50

30.39

12.43

19.94

33.88

8.38

13.42

23.40

8.90

ST-ABC

19.61

30.76

13.58

21.88

34.19

9.55

14.97

23.77

10.29

ASTTN

18.51

30.20

12.21

20.05

32.94

7.94

15.07

24.10

8.60

TSGDC

18.80

31.08

12.67

19.96

33.25

8.54

14.12

23.39

9.63

PDFormer

18.32

29.97

12.10

19.83

32.87

8.53

13.58

23.51

9.05

IEEAformer

18.22

30.31

11.99

19.11

32.60

8.01

13.49

23.20

8.89

TSTGNN

19.06

30.52

12.72

20.40

33.83

8.72

15.69

24.72

9.86

TDMGCN

18.18

30.32

12.02

18.66

32.01

7.92

12.86

22.73

8.94