Abstract:

 In terms of item counts and income, high-utility itemset mining (HUIM) has emerged as a significant step in the pattern mining process. A number of approaches for mining high-utility itemsets have been developed (HUIs). Because these algorithms usually produce a significant number of detected patterns, a more compact and lossless version has been devised. The newly described closed high utility itemset mining (CHUIM) algorithms were designed to work with certain datasets (e.g., those without probabilities). Real-world data bases may include items or itemsets associated with probability values. Several methods for rapidly mining frequent patterns from uncertain data bases have been developed, however there is no technique to mine CHUIs from this kind of database. CPHUI-List is a new and efficient method for mining closed prospective high-utility itemsets (CPHUIs) from uncertain datasets without developing candidates, as described in this study. The proposed method is based on DFS and takes use of non-CPHUIs' downward closure characteristic, as well as a high transaction-weighted probabilistic mining top run. The experiment findings show that the proposed technique beats the CHUI Miner in terms of execution time and memory use.

Keywords: