Density-based clustering algorithm: DBSCAN
- It requires two parameters are predefined: "radius of influence" and "Threshold Effect"
- Advantages are: kmeans not like the previously determined as the value of k, accommodate nonconventional data, can be used to find outliers (outliers)
- The disadvantages are: two parameters still need to be determined in advance
- The key to understanding: the development of similar MLM downline or tribal expansion
- Schematic:
- DBSCAN visual display