¶ Description
DBSCAN Clustering.
¶ Parameters
¶ Parameters tab
Parameters:
- Input table partitioning (1st pin)
- Variables to include in clusters
- Plot distance screen
- ε value
- Number of minimum points in the ε region
- Borderpoints
- Search strategy
- Split rule
- A short description
¶ Description tab
Parameters:
- Script name
- Short description
- Revision
- Decription
¶ Configuration tab
See dedicated page for more information.
¶ About
This action is mainly for explanatory/teaching purposes. If you want to create a better segmentation, you should use Stardust.
The DBScan algorithm is often used in geographic settings, to understand segments about mobility, for example.
It is performed in two steps. First, run to plot the distance plot, and draw and imaginary line between the last point of inflection, and the vertical axis. This will tell you roughly what the value of Epsilon should be (see red line, in this case, about 15)
You will then be able to get segment assignment and a plot:
