Usage:

python ai_playground.py
# This will create a tree based on params listed in the main function.
# Then based only on leaves, will attempt to infer the tags at each node in the tree.
# currently this seems to be able to split a group of leaves into two clearly defined
# groups until depth 3 (where there are 4 leaves left to split)


TODO:
* Figure out bug defined above.
* Print original tree and compare it with the newly inferred tree.
* Fix the deletion process to choose random starting index to delete (instead of just at end)
* Save randomly generated tree into .pkl file (cPickle).
* Save result into .pkl file (cPickle).
* Add options for parameters.

Some tree view tool: http://etetoolkit.org/treeview/