The Mining Algorithm of Maximum Frequent Itemsets Based on Frequent Pattern Tree

Table 4

Pseudocode of FP-MFIA.

---

Algorithm 3: The algorithm of FP-MFIA.

---

Input: The FP-tree of transaction database D, the Header table of frequent item, the minimum support count s, the frequent item list L_DF in D

Output: The MFI-list

Process:

MFI-list = Ø;

Preorder traversal of the FP-tree to obtain a simplified FP-tree, and initialize the MFI-list at the same time;

Traverse the MFI-list and simplify the FP-tree;

p = Header table[| Header table|]; //Analyze each node in the Header table in reverse order, p first points to the last node;

while(p exists){

p = p- > item_links; //p points to the first node in the FP-tree with the same item name

q = q- > first_MFI; //q points to the MFI-list node with the same item name as p in the MFI-list

while (q exists)

{According to p- > item_links and the node_links of the nodes in the FP-tree, find all the nodes nd₁, nd₂,...,nd_h with the item name p- > item_name;

for(i = 1; i ≤ h; i++)

for(j = 1; j ≤ h; j++)

If() then

{Convert nd_i- > node_pre, nd_j- > node_pre to bit vector;

nd = nd_i;

n = nd- > node_pre∧nd_j- > node_pre;

if(the number of “1” in nd > 1) then

{nd.count = nd_i- > node_count + nd_i- > node_count;//Calculate the support count for nd

q = q- > first_MFI;

If(nd.count ≥ s)

{while(q- > next_MFI)

if(nd is not a subset of an element in the itemset pointed to by q）q = q- > next_MFI;

else break; }

If(q- > next_MFI = = null) q- > next_MFI = nd; }

}}}}

---