Data Mining essay 2 Clustering algorithm

Clustering (cluster)

Partitioning Methods:

K-means

Saquential Leader

Model Based Methods

Density Based Methods

Hierachical Methods

Unsupervised learning (unsupervised learning)

No Labels

Data driven

Special needs to be aware of:

Arbitrary shape handles a wide variety of shapes noise and outlier noise, and the ability to process outliers

---------------------------------------------------------------------

K-means algorithm

Evaluation

This note is mainly about the effect of spherical data is better, for other effects may be problematic

Silhoustte

Data Mining essay 2 Clustering algorithm