peterwittek / somoclu

Massively parallel self-organizing maps: accelerate training on multicore CPUs, GPUs, and clusters
https://peterwittek.github.io/somoclu/
MIT License
266 stars 69 forks source link

umatrix to #121

Closed Tortoise17 closed 5 years ago

Tortoise17 commented 6 years ago

Dear Colleagues.

Good Morning. I have question. after training in SOMOCLU, I get Umatrix as the lets supposr 10 / 10 cluster. If i have to get the specific cluster only and use that to get the BMUS, is there any way to segregate? like I have seen the zoom example. but if I can get the threshold value and make the use of it as classified neurons for further use.

because Umatrix has 10 / 10 matrix and how there are neurons information is sticked internally and get back the cluster. If you can guide me. i want to separate a part of cluster and use that clustered part for further mapping.

Thank you.