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

Table 3

Pseudocode for constructing MFI-list.

---

Algorithm 2: Algorithm of constructing MFI-list.

---

Input: FP-tree

Output: MFI-list

Process:

current = T://current points to the current node to be traversed, and its initial value points to the tree root T

InitStack(s); //initialize stack s

There are untraversed paths in the FP-tree

{while (current- > node_count ≥ min_sup)

{current points to a child node that has not been visited;

Push(s,current- > node_name);

}

//When there is no child node or the node_count value of the child node is less than min_sup

MFI = {collection of items in stack s};

Convert the MFI into the corresponding bit vector and link it to the corresponding maximum frequent itemset linked list

current- > node_pre = MFI;

While(current has child nodes in the current path)

{pre = current;

current points to its child node;

current- > node_pre = pre- > node_pre∪current- > node_name;//Take the frequent itemset composed of the union of the node_pre of its parent node and the node_name of the node

}}

---