Biomedical Engineering Reference
In-Depth Information
Equation 1.
2
−
b
∆
t
∆
t
x
(
t
+
∆
t
−
)
x
(
t
−
)
x
(
t
+
∆
t
)
=
x
(
t
)
+
+
2
+
b
∆
t
c
c
2
+
b
∆
t
1
+
x
(
t
+
∆
t
−
)
1
+
x
(
t
−
)
Figure 16. Comparison between real values of the time series and the RANN predictions (3 hidden
neurons)
Table 6. Values of the connection weights of the RANN and the line gradients of the time decreased
activation
NETWORK CONNECTIONWEIGHTS
Neuron
1
2
3
4
5
1
-1.682
0.992
-1.101
1.606
5.965
9.932
2
2.866
1.429
-4.339
0.044
3
-7.999
4.963
5.398
-2.304
9.843
4
0.765
2.160
7.411
-9.971
9.991
5
0.572
-1.069
0.011
-9.989
-0.564
LINE GRADIENTS
Neuron
1
2
3
4
5
1
-0.347
-1.000
-0.146
-0.089
-0.197
2
-0.776
-0.096
-0.103
-0.597
-0.031
3
-0.999
-0.154
-0.929
-0.979
-0.985
4
-0.614
-0.472
-0.446
-0.236
-0.156
5
-0.096
-0.132
-1.000
-0.451
-1.000
Search WWH ::
Custom Search