pyclustering icon indicating copy to clipboard operation
pyclustering copied to clipboard

[pyclustering.cluster.agglomerative] Ward's method for cluster merging

Open annoviko opened this issue 7 years ago • 0 comments

Introduction The was a topic on stackoverflow referred to pyclustering where Ward's method was mentioned. It is a method that is used as criterion for choosing pair of clusters for merging (as well as existed in our implementation SINGLE_LINK, COMPLETE_LINK, etc.). Ward's method should be introduced to agglomerative clustering algorithm.

Resources

  • Wiki page: https://en.wikipedia.org/wiki/Ward%27s_method
  • Ward, J. H., Jr. (1963), "Hierarchical Grouping to Optimize an Objective Function", Journal of the American Statistical Association, 58, 236–244.

Task description

  1. The method should be implemented for 'pyclustering.cluster.agglomerative' module. Existed 'type_link' (that consists of SINGLE_LINK, COMPLETE_LINK, AVERAGE_LINK, CENTROID_LINK) should be renamed to much appropriate enumerator.
  2. Ward's method should be introduced in the renamed enumerator.
  3. Ward's method should be implemented.
  4. Unit-tests should be implemented for this criterion to 'pyclustering/cluster/tests/unit/ut_agglomerative.py'.
  5. Examples should be implemented for this criterion to 'pyclustering/cluster/examples/agglomerative_examples.py'.

annoviko avatar Aug 29 '17 13:08 annoviko