Information Technology Reference
In-Depth Information
(a)
1.1
1.05
1
0.95
0.9
0.85
0.8
0.75
0.7
0.65
0.6
0
10
20
30
40
50
60
70
80
90
100
time (s)
(b)
1.2
1.1
1
0.9
0.8
0.7
0.6
0.5
0.4
0
10
20
30
40
50
60
70
80
90
100
time (s)
(c)
8
6
4
2
0
-2
0
10
20
30
40
50
60
70
80
90
100
time (s)
Fig. 7 Evolution of control and power alteration parameters in GFS and output power for two
compared approaches. a Controlled torque. b Output aerodynamic power of wind turbine in Pu.
c Pitch angle signal for GFS
regard the trained neural network as an opaque and aim to extract rules that map
inputs directly into outputs.
The calculative complexity of extracting rules from trained neural networks and
the complexity of extracting the rules straight from the data are both NPhard (Golea
1996
), thus many of these cases contain a salient theoretical discovery in this area.
Roy intelligently made known the difference between the idea of rule extraction and
traditional connectionism (Roy
2000
).
There are two theory intentions in Rule Extraction from trained neural network.
The ANNs or weights are evaluated to extract the whole rule set witch should
illustrate he complete wisdom.
Search WWH ::
Custom Search