para.config

数据挖掘经典的hierarchial clustering algorithm

                                For clustering of all projections of attributes:

CorD                            0(clustering analysis) or 1(density analysis) *
TotalMemSize in bytes           about 5% of dataset size for each projection *
TotalBufferSize in bytes        0% of TotalMemSize
TotalQueueSize in bytes         0% of TotalMemSize
TotalOutlierTreeSize in bytes   0% of TotalMemSize

                                For clustering of each projection of attributes:
                                the following set of parameters need to be set.
                                Please make sure the parameters relevant to
                                the projection dimension are set correctly.

WMflag                          0 or 1  weight/moving or not *
W vector                        (relevant to dimension)  weight vector *
M vector                        (relevant to dimension)  moving vector *

PageSize in bytes               1024

BDtype                          2
Ftype                           0
Phase1Scheme:                   1
RebuiltAlg:                     0
StatTimes                       3
NoiseRate                       0.25

Range                           2000

CFDistr                         1 (only for density analysis)
H                               0 (only for density analysis)
Bars on each dimension          (relevant to dimension), for density analysis *

K                               number of clusters, for clustering analysis *
InitFt                          0.0
Ft                              0.0, for clustering analysis
Gtype                           1, for clustering analysis
GDtype                          2, for clustering analysis
Qtype                           0
RefineAlg                       1, for clustering analysis
NoiseFlag                       0, for clustering analysis
MaxRPass                        1, for clustering analysis *