|
1. using the large set of items Li – 1 found in the i –1 traversal and the Apriori-gen function to generate the candidate set Ci; and 2. Scan the database and calculate the support of Ci. The Apriori algorithm is described as follows. L=find_frequent_1-itemsets(D);l/All frequent 1-item sets found for (i=2: L;‵.≠;i++) { C;=apriori_gen(L;.,minsupp);ll Generating candidate sets for each transaction t D;//Scan all transactions { C=subset(Ci,t);//Identify the set of all candidates belonging to t for each candidate c C c.count++;l/Support count increments of 1 } Li={cCk c.countminsupp};/ Extracting frequent i-item sets } return L=UkLk;l/ |