Efficiently Identifying Exploratory Rules’ Significance - Data Mining: Theory, Methodology, Techniques, and Applications - page 75

the rule antecedents from 3 to 8, and then run the program with no limit on the
maximum number of conditions allowed on rule antecedents. After this, the dif-
ference set statistics derivation approach and the circular intersection approach
are introduced respectively, before the ecient algorithm in table 2 is ran fol-
lowing the same procedure. For ipmus.la.98, ipmus.la.99, ticdata2000, census
income and covtype , which are relatively larger databases, we only ran the pro-
grams with maximum number of conditions allowed on rule antecedents set to
3, 4, and 5. We plot the allowed number of maximum conditions on antecedents

(a) (b)

(c) (d)

(e)

Fig. 5. Comparison of Running Time before and after applying data access saving
techniques for (a) Ipums.la.98 ,(b) ipums.la.99 ,(c) Ticdata2000 ,(d) Census income ,
and (e) covtype

with maximum number of conditions allowed on rule antecedent set to

3, 4 and 5