Table 2 Performance comparison of WMF-Traffic with state-of-the-art baselines across four traffic prediction datasets. Results are reported as mean ± standard deviation over 5 independent runs. Best results are highlighted.

From: Multi-scale Wavelet-Mamba framework for spatiotemporal traffic forecasting

Method

METR-LA

PEMS-BAY

PEMS04

PEMS08

\(\operatorname {MAE}\downarrow\)

\(\operatorname {RMSE}\downarrow\)

\(\operatorname {MAPE}\downarrow\)

\(\operatorname {Avg.}\)

\(\operatorname {MAE}\downarrow\)

\(\operatorname {RMSE}\downarrow\)

\(\operatorname {MAPE}\downarrow\)

\(\operatorname {Avg.}\)

\(\operatorname {MAE}\downarrow\)

\(\operatorname {RMSE}\downarrow\)

\(\operatorname {MAPE}\downarrow\)

\(\operatorname {Avg.}\)

\(\operatorname {MAE}\downarrow\)

\(\operatorname {RMSE}\downarrow\)

\(\operatorname {MAPE}\downarrow\)

\(\operatorname {Avg.}\)

\(\operatorname {VAR}\)

4.52±0.23

8.34±0.35

13.87±0.51

8.91

2.56±0.17

5.78±0.31

6.21±0.26

4.85

30.45±0.72

47.23±0.96

22.67±0.64

33.45

24.89±0.58

39.56±0.78

17.89±0.52

27.45

\(\operatorname {SVR}\)

4.28±0.21

8.12±0.33

12.45±0.46

8.28

2.34±0.15

5.48±0.28

5.76±0.23

4.53

29.34±0.68

45.67±0.91

21.34±0.59

32.12

23.78±0.54

38.12±0.73

16.78±0.48

26.23

\(\operatorname {LSTM}\)

3.98±0.19

7.76±0.29

11.23±0.39

7.66

2.09±0.13

5.01±0.24

5.14±0.19

4.08

27.89±0.61

43.56±0.83

19.78±0.53

30.41

22.34±0.49

36.78±0.68

15.67±0.43

24.93

\(\operatorname {GRU}\)

3.84±0.18

7.52±0.27

10.76±0.36

7.37

1.96±0.12

4.78±0.22

4.89±0.17

3.88

27.12±0.57

42.78±0.79

19.12±0.49

29.67

21.89±0.46

36.23±0.64

15.23±0.41

24.45

\(\operatorname {Transformer}\)

3.76±0.16

7.38±0.25

10.45±0.33

7.20

1.87±0.11

4.61±0.20

4.67±0.16

3.72

26.45±0.53

41.89±0.75

18.56±0.46

28.97

21.34±0.43

35.67±0.61

14.89±0.38

23.97

\(\operatorname {Informer}\)

3.69±0.17

7.25±0.26

10.21±0.34

7.05

1.82±0.12

4.48±0.21

4.53±0.17

3.61

26.01±0.54

41.23±0.77

18.23±0.47

28.49

20.98±0.44

35.23±0.62

14.61±0.39

23.61

\(\operatorname {Autoformer}\)

3.62±0.15

7.14±0.24

10.01±0.31

6.92

1.76±0.11

4.35±0.19

4.42±0.15

3.51

25.45±0.51

40.34±0.72

17.78±0.44

27.86

20.56±0.41

34.67±0.58

14.28±0.36

23.17

\(\operatorname {DCRNN}\)

3.67±0.16

7.21±0.25

10.15±0.32

7.01

1.79±0.11

4.42±0.20

4.48±0.16

3.56

25.78±0.52

40.89±0.74

18.01±0.45

28.23

20.87±0.42

35.01±0.59

14.45±0.37

23.44

\(\operatorname {GraphWaveNet}\)

3.64±0.15

7.18±0.23

10.08±0.30

6.97

1.77±0.10

4.38±0.18

4.45±0.14

3.53

25.61±0.49

40.56±0.71

17.89±0.43

28.02

20.71±0.40

34.78±0.57

14.34±0.35

23.28

\(\operatorname {ASTGCN}\)

3.59±0.14

7.09±0.22

9.95±0.29

6.88

1.74±0.10

4.31±0.17

4.38±0.13

3.48

25.34±0.47

40.12±0.68

17.67±0.41

27.71

20.45±0.38

34.45±0.55

14.18±0.33

23.03

\(\operatorname {STGODE}\)

3.56±0.13

7.03±0.21

9.87±0.27

6.82

1.71±0.09

4.25±0.16

4.32±0.12

3.43

25.12±0.45

39.78±0.65

17.45±0.39

27.45

20.23±0.36

34.12±0.52

13.98±0.31

22.78

\(\operatorname {S4}\)

3.57±0.14

7.05±0.22

9.91±0.28

6.84

1.72±0.10

4.27±0.17

4.35±0.13

3.45

25.23±0.46

39.89±0.67

17.56±0.40

27.56

20.31±0.37

34.23±0.53

14.05±0.32

22.86

\(\operatorname {PatchTST}\)

3.54±0.13

7.01±0.20

9.84±0.26

6.80

1.70±0.09

4.23±0.16

4.29±0.12

3.41

25.05±0.43

39.65±0.63

17.38±0.38

27.36

20.15±0.35

34.01±0.51

13.92±0.30

22.69

\(\operatorname {Mamba}\)

3.51±0.12

6.97±0.19

9.78±0.25

6.75

1.68±0.08

4.19±0.15

4.25±0.11

3.37

24.89±0.41

39.45±0.61

17.23±0.36

27.19

19.98±0.33

33.78±0.49

13.81±0.28

22.52

WMF-Traffic (Ours)

3.05±0.09

6.12±0.16

8.54±0.21

5.90

1.45±0.07

3.67±0.13

3.72±0.10

2.95

21.56±0.37

34.78±0.56

15.12±0.33

23.82

17.23±0.30

29.45±0.45

12.01±0.26

19.56