238
G. Bai et al.
Fig. 21.1 Flowchart of
pattern recognition
Therefore, the total number of combinations is R = N × L. Then we can get the
R parameter combinations through the traversal.
All the parameter combinations and their IDs can be obtained by the above method,
then a prior k-means algorithm as the following was used to cluster.
1. Initialization: calculate the number of clusters according to the maximumminimum distance criterion [16], and define the obtained K center as
C 10 ,C 20 ,…,C K0 .
G. Bai et al.
Fig. 21.1 Flowchart of
pattern recognition
Therefore, the total number of combinations is R = N × L. Then we can get the
R parameter combinations through the traversal.
All the parameter combinations and their IDs can be obtained by the above method,
then a prior k-means algorithm as the following was used to cluster.
1. Initialization: calculate the number of clusters according to the maximumminimum distance criterion [16], and define the obtained K center as
C 10 ,C 20 ,…,C K0 .