Dear researchers,
I am using SOM to cluster my data in python 3.6 and I have get the result visually through various maps. My question is how to obtain the results (description) of clustering process such as: members of each output cluster, the distances between these clusters? This is required to check the coherence factor inside each cluster and the level of dissimilarity between the overall clusters.
Thanks
Multiple implementations of SOM in Python are available and I have used one of my own.(http://damayanthiherath.github.io/selforganisingmaps/).
Nevertheless, I believe you are referring to this implementation :
https://pypi.python.org/pypi/sompy/0.1.1
which has the source at https://github.com/ttlg/sompy. If you look at the script,s the output map provides you the weight vectors of the each node on the map.
In this case, generated self organising map can only be used to visualise the input data in the reduced dimensionality (2D). With the weight vectors in hand, you can implement a function to see where do the original inputs fall on the map. It is done by placing each input on the node with which it has the least distance (in this case considering the Euclidean distance).
OK.. I understand that I should implement my own function... Does yours solve this issue?
However, it is the most accurate answer to my question.. So many thanks..
No, Mine does not include a function for it as well. However, an example of mapping input is in
https://github.com/damayanthiHerath/selforganisingmaps/blob/master/somTestIris.py.
- Badges
- Science topic
- Similar topics
- Computer Science and Engineering
- Programming Languages
- Python