Documenti di Didattica
Documenti di Professioni
Documenti di Cultura
Note: insert_tree([p|P], T) is
performed as follows: Analysis
IF T has a child N such that Two scans of the DB are necessary:
N.item_name=p.item_name, then the first collects the set of frequent
increment N's count by 1 items and the second constructs the
E L S E create a new node N, and let its F P-tree.
count by 1, its parent link be linked to T,
and its node-link be linked to the nodes T he cost of inserting a transaction
with the same item_name via the node- T rans into the FP-tree is
link structure O(|Trans|), where | T rans| is the
IF P is nonempty, call insert_tree(P,N) number of frequent items in T rans.
recursively
F P-growth: the FP-tree-based
mining method
F P-tree contains the complete Starts from a frequent length-1
information for frequent pattern mining.
pattern
T he size of the FP-tree is bounded by the
size of the database, but due to frequent Examines only its conditional
items sharing, the size of the tree is pattern base
usually much smaller than its original
database. Constructs its FP-tree
High compaction is achieved by placing Performs mining recursively on
more frequently items closer to the root the tree
(being thus more likely to be shared).