Table 3 Performance comparison of ablation experiment models.

From: Research on deep learning model for stock prediction by integrating frequency domain and time series features

Model

NASDAQ

NYSE

 

IC

RIC

Prec@N

SR

IC

RIC

Prec@N

SR

Original

41.23

41.27

40.79

40.65

49.67

43.77

41.34

43.22

w/o ATFNet

39.85

40.03

38.91

39.12

47.93

41.12

39.30

40.03

w/o Mixer

39.17

39.40

38.15

38.42

47.18

40.65

38.20

39.21

Only ATFNet

37.62

37.88

36.73

36.85

46.02

39.12

37.14

37.95

Replace Fourier

40.31

40.54

39.85

40.00

48.70

42.53

40.48

41.42

w/o StockMixer

38.25

38.41

37.01

37.32

46.73

40.00

38.16

39.01

w/o Conv

40.42

40.48

39.76

39.89

48.96

42.90

40.87

41.80

Drop

35.54

35.71

34.63

34.80

44.67

37.05

35.22

35.81

Back to article page