Gene clustering: 2001:
2
Finding Rules: Algorithm
§
For each instance NOT in cluster:
•
find set of words in cluster but not in instance
§
Collect singleton sets
(and remove supersets)
•
= “necessary” words
§
Collect all other words in sets
•
= “sufficient” words
§
Collect words in cluster but not in any set
•
= “supplementary” words
§
Convert sets from CNF to DNF