Filtering decision rules using generators and closed itemsets

Apriori algorithm can generate huge sets of decision rules. A lot of these rules, especially in multilevel mul-
tidimensional environment, are redundant or useless. To reduce this unwanted eect the author proposes to
utilise generators or closed itemsets, originally used in lossless itemsets representations, to lter decision rules
during their generation. These two modications of the Apriori algorithm enable reduction of a size of resultant
decision rules' sets without lost of information about relations in training data. Choice of the proposed modica-
tions depends on user's preference about length of the rules. Filtering using generators gives shorter rules while
ltering with closed itemsets results in longer rules. Both modications were evaluated on large sets of training
data from faults simulations experiments.