|
|
The TwoStep node uses a two-step clustering method. The first
step makes a single pass through the data to compress the raw input
data into a manageable set of subclusters. The second step uses a
hierarchical clustering method to progressively merge the subclusters
into larger and larger clusters. TwoStep has the advantage of automatically
estimating the optimal number of clusters for the training data. It
can handle mixed field types and large data sets efficiently. |
Example
node = stream.create("twostep", "My node")
node.setPropertyValue("custom_fields", True)
node.setPropertyValue("inputs", ["Age", "K", "Na", "BP"])
node.setPropertyValue("partition", "Test")
node.setPropertyValue("use_model_name", False)
node.setPropertyValue("model_name", "TwoStep_Drug")
node.setPropertyValue("use_partitioned_data", True)
node.setPropertyValue("exclude_outliers", True)
node.setPropertyValue("cluster_label", "String")
node.setPropertyValue("label_prefix", "TwoStep_")
node.setPropertyValue("cluster_num_auto", False)
node.setPropertyValue("max_num_clusters", 9)
node.setPropertyValue("min_num_clusters", 3)
node.setPropertyValue("num_clusters", 7)
Table 1. twostepnode properties| twostepnode Properties |
Values |
Property description |
|---|
| inputs |
[field1 ... fieldN] |
TwoStep models use a list of input fields, but no target. Weight
and frequency fields are not recognized. See the topic Common Modeling Node Properties for more information. |
| standardize |
flag |
|
| exclude_outliers |
flag |
|
| percentage |
number |
|
| cluster_num_auto |
flag |
|
| min_num_clusters |
number |
|
| max_num_clusters |
number |
|
| num_clusters |
number |
|
| cluster_label |
String
Number |
|
| label_prefix |
string |
|
| distance_measure |
Euclidean
Loglikelihood |
|
| clustering_criterion |
AIC
BIC |
|
