Comparative analysis of high average-utility patterns algorithms

Utility mining is a current emerging field in data mining. Utility mining raises various forms. They are high utility itemset mining, utility frequent itemset mining, negative utility itemsets mining, rare high utility itemset mining etc. All these itemset mining does not consider the size of itemsets. A recent development in utility mining is high average-utility itemset mining. This average utility mining considers length of itemsets along with the utility of itemsets. There are several algorithms proposed to retrieve high average utility itemsets in a database. The main objective of this work is to compare three high average-utility patterns algorithms: high average-utility patterns (HAUP) algorithm, high average-utility itemset-miner (HAUI-Miner) algorithm and efficient high average-utility pattern mining (EHAUPM) algorithm. The execution time and memory space are considered as performance measures for the comparison. It was found that EHAUPM algorithm performs better than other algorithms.