Skip to main content

Advances, Systems and Applications

Table 4 Test RMSE and MAE for different methods

From: A hybrid attention and time series network for enterprise sales forecasting under digital management and edge computing

Method

ESDA

ESDB

ESDC

RMSE

MAE

RMSE

MAE

RMSE

MAE

CNN

53.8 ± 3.6

47.9 ± 3.1

49.9 ± 3.8

41.5 ± 3.3

47.2 ± 2.9

39.6 ± 2.6

LSTM

49.5 ± 2.7

41.3 ± 2.5

46.1 ± 1.9

36.5 ± 2.7

43.5 ± 3.3

34.1 ± 3.0

GRU

46.7 ± 4.1

35.2 ± 2.7

43.2 ± 2.8

32.8 ± 5.6

39.2 ± 3.8

30.5 ± 2.8

DEEPAR

51.2 ± 2.6

44.6 ± 2.3

47.9 ± 2.2

38.7 ± 4.3

45.5 ± 3.2

37.2 ± 3.2

Nbeats

48.1 ± 2.9

37.9 ± 3.3

44.6 ± 2.5

33.9 ± 3.7

42.1 ± 2.8

32.3 ± 3.5

edRVFL

45.1 ± 3.6

32.6 ± 3.7

43.6 ± 3.5

32.1 ± 2.9

38.9 ± 4.4

28.2 ± 4.1

DESN

47.3 ± 1.9

37.1 ± 1.7

44.9 ± 5.1

35.5 ± 3.4

42.1 ± 4.1

32.6 ± 2.7

BiLSTM

44.5 ± 1.5

31.7 ± 1.8

41.4 ± 2.7

29.7 ± 2.4

36.8 ± 2.9

25.4 ± 2.2

HATT-MSCNN-IBiLSTM

37.6 ± 1.1

23.5 ± 1.3

31.6 ± 2.0

20.2 ± 1.6

25.7 ± 1.9

17.3 ± 0.8