Table 4 The performance of different wavelet basis functions on four data sets. Sym4 Db Coif Haar SSEC 65.14

Expert Systems With Applications 200 (2022) 116941
5
advantages of WT is that it decomposes a signal according to the frequency and represents the signal in the time domain. As for the wavelet transformation, both the time and the frequency information from the signal are retained. It is thus a more powerful transformation for time-
–
frequency analysis (
Rhif et alb. WT can decompose signals on different timescales. The purpose of WT is to translate and shrink the wavelet basis function. The wavelet basis function is defined as
ψ
a,b
(
t) =
1
̅̅̅
a
√
ψ
(
t − b
a
),
a > 0, bi R.
(4) where a is a scale parameter and b is a time center parameter. When a
and b change continuously, the wavelet transform process is continuous wavelet transform (CWT. Fora signal f(t), CWT is defined as
W
f
(
a, b) =
∫
∞
−
∞
f (t)
ψ
*
a,b
(
t)dt =
1
̅̅̅
a
√
∫
∞
−
∞
ψ
*
a,b
(
t − b
a
)
f (t)dt
(5)
Table 6
The model performance of different forecast target in the SSEC data set. forecast target Accuracy Precision Recall F score
0 0.643 0.641 0.641 0.641 1
0.700 0.701 0.701 0.701 2
0.714 0.712 0.712 0.712 3
0.733 0.739 0.734 0.735 4
0.731 0.729 0.731 0.731 5
0.728 0.738 0.729 0.729 6
0.735 0.744 0.734 0.736 7
0.736 0.741 0.736 0.736 8
0.735 0.740 0.736 0.736 9
0.736 0.742 0.737 0.739

Download 2.42 Mb.

Share with your friends: