算法2:准反常项集挖掘算法

输入:数据集D的垂直数据表示形式 D vert [ P ] ,提升度阈值 λ ,更正操作数量k

输出:准反常项集QAbNI

1

procedure QAbNIMiner ( D vert [ P ] , P I , λ )

2

QAbNI

3

for all m I occurring in D vert [ P ] in reverse order do//depth-first search

4

N P { n } //expanding the itemset N

5

if k | D | ( λ 1 2 1 ) then

6

if not isGenerator(N) then

7

continue

8

storeGenerator(N)

9

if | N | > 1 and wSup ( N , D clean ) λ 2 λ then

10

if a subset of N has been pruned then

11

continue

12

if mpwHlift ( N , D ) < λ then

13

QAbNI QAbNI { N }

14

If 1 λ > min S N { wSupCnt ( S , D clean ) + k wSupCnt ( N \ S , D clean ) + k } then