1. In a transaction database 10 items (i=1,...,10) occur independetly of each other with probability p_i=1/i. 
a) Which itemsets are going to be frequent if we apply a relative frequency threshold of 1%?
b) Which association rules are going to be of interest?

2. Suppose our transaction database contains 100 different products and 100 market baskets. Item i can be found in basket b if i divides b (without a remainder).
a) What is going to be the sum of all items over all the baskets?
b) Which items are going to be frequent if we apply a support threshold of 5?
c) Which item pairs are going to be frequent if we apply a support threshold of 5?
d) Which (arbitrary sized) item sets are going to be frequent if we apply a support threshold of 5?
e) What is going to be the confidence and interestingness of the association rules {5,7}->2 and {2,3,4}->5?
f) Which bucket(s) will have the most items in it? Write an Octave function that calculates the number of items in each basket!
g) How can those association rules of form A->B characterzized which has a confidence score 1,0?

3. Find 3-frequent item sets from the lecture notes.