Certain investigations on mining frequent and high average utility itemsets using meta heuristics algorithms

Abstract

newline Data mining is a field of discovering interesting patterns, associations, newlinetrends and useful insight within large datasets. The main objective is to mine newlinevaluable information which will help in making informed decisions. Frequent newlineitemset mining (FIM) is a technique employed in data mining to uncover newlinefrequent patterns that occur in a dataset. This will mine frequent patterns, newlinepatterns which have support value greater than or equal to the minimum newlinesupport value fixed by the user. Support value determines the number of times newlinethe pattern has appeared in the dataset. But this does not highlight the true newlineimportance or value of the items within the patterns. Utility mining (UM) newlineovercomes the drawback of FIM and it considers the utility or importance of newlineeach item in the dataset rather than frequency of occurrence. The key newlineobjective of UM is to mine High Utility Itemsets (HUIs), itemsets with utility newlinevalues greater than the user-defined minimum utility values. Often, the newlineitemsets which are length will have high utility values since it has a greater newlinenumber of items in it. The reason behind this drawback is, the length of the newlineitemsets is not considered as a metric.