Database Reference
In-Depth Information
Normal
Cyclic
70
70
60
60
50
50
40
40
30
30
20
20
10
10
0
0
0
10
20
30
40
50
60
0
10
20
30
40
50
60
Increasing
Decreasing
70
70
60
60
50
50
40
40
30
30
20
20
10
10
0
0
0
10
20
30
40
50
60
0
10
20
30
40
50
60
Upward
Downward
70
70
60
60
50
50
40
40
30
30
20
20
10
10
0
0
0
10
20
30
40
50
60
0
10
20
30
40
50
60
Fig. 5. Examples of the
Control
data set. Two examples of the same class are shown
in each graph.
The results are shown in Figure 8 and in Table 9. For this data set,
boosting point based literals is not able to learn anything useful, the error
is always greater than 70%. Nevertheless, using boosting of interval based
literals and more than 80 literals, the error is less than 1%. For early classi-
fication, the error is under 1% with the 80% of the series length. The result
reported in [Roverso (2000)], using recurrent neural networks and wavelets
is an error of 1.4%, but 4.5% of the examples are not assigned to any class.
