Pattern mining: itemset enumeration with borders

What the exam asks you to produce (archetype A25): the queue (the session 10 solutions call it the stack) written out at every step, the frequency you used beside every pop, the itemsets output in order, and the S and G sets at the end.

generation rule new itemsets go to refine when max threshold boundary lattice edges predict

History: one row per step, the shape of a correct written answer

FAI exercise exam only: absolute counts, never support percentages, never confidence, never an Apriori self join. Those belong to the Machine Learning exam on 20 August. Here a threshold is a number of transactions.