TO DO:

*  run experiments based on classification accuracy
*  read committee clustering paper
*  implement committee clustering
*  develop metrics for clustering accuracy
*  look into sparse arff
*  integrate Andrian's LSI code
*  implement twncb

IN PROGRESS:

o  tweak genic removal criteria 
	currently: deletes clusters with less than 2 members

DONE:
X  upgrade genic to support nominal attributes [done]
X  modify clustering distance metrics to support nominal attributes [done]
X  fix bugs in kmeans [done]