Consider the following transaction dataset T1 a d e T2 a b c

Consider the following transaction dataset. T1: a, d, e T2: a, b, c, e T2: a, b, d, e T4: a, c, d, e T5: b, c, e T6: b, d, e T7:c, d T8: a, b, d a) Compute the support for itemsets {e}, {b, d}, and {b, d, e}. b) Compute the confidence for the association rules {b, d} rightarrow {e} and {e} rightarrow {b, d). c) Is confidence a symmetric measure? d) List all 3-itemsets with minimum support 0.2.

Solution

a)Support is \'the number of times {e} is in all the transactions / total number of transaction\' multiplied by 100.

Support of {e} = 6/8 =3/4 = 0.75 = 75%

Support of {b,d} = 3/8 = 0.375 = 37%

Support of {b,d,e} = 2/8 = 0.25 = 25%

b)for example X->Y is an association rule the confidence is computed is as follows

Confidence is \'number of times X and Y is repeated in all the transactions / number of times X is repeated\'.

therefore {b.d} -> {e} = 2/6 = 1/3 = 0.33 = 33%

{e} -> {b.d} = 2/3 = 0.66 = 66%

Confidence is not a symmetric measure

d) these are the three item set list which supports minimum support

{a,b,d} = 0.25

{b,c,e} = 0.25