The words contained in this file might help you see if this file matches what you are looking for:

...Lecture notes a advanced data mining tddd clustering and association analysis jose m pena ida linkoping university sweden correctness of the apriori algorithm proof is not unique you can nd one in article by agrawal srikant available from course website our own alternative be found below weprove induction on k that correct we prove result for then under assumption up to combining this two facts conclude any first recall d minsup input transactional database minimum support output all large itemsets l do ck gen lk generate candidate t c such count cksc return klk trivial case line hypothesis assume now it suces because lines simply frequency candidates thus nothing go wrong there function superset self join i j if call genrules am minconf contradiction rule generation contrary missed let denote missing rules with largest antecedent note wrongly implies has outputted since as proven previous section cannot have when konly called evaluated therefore must subsequent calls e...