hello respected Members!
i am working with dynamic features at my input whose value are changing per time slot of agent movement. due to dynamic nature i want to do partition by using K-mean clustering. here my original state shape reduce but here i am facing a problem i dunnu how can i convert this kmean to hashcode possible Anchor graph mapping, bcos this can map the data size to a fix length. then this hash code i want to integrate with my input state. further if anyone have understanding on this, i will give him/her a paper related to problem.
keep note:' i dun want to share my code publically if anyone interested they should do work on my pc