Mining high utility partial periodic pattern by GPA
Tzung‐Pei Hong, Jen-Hao Hsu, Kung-Jiuan Yang, Guo-Cheng Lan, Jerry Chun‐Wei Lin, Shyue-Liang Wang · 2017
Different from full periodic patterns, partial periodic patterns could ignore the occurrence of some events in time positions. In this paper, we have presented a gradually pruning algorithm (GPA) for reducing the number of candidate patterns in the mining process. It is based on the two-phased periodic utility upper-bound (PUUB) model and could avoid information loss. Compared to the original approach without gradually pruning, the one proposed here could reduce the execution time but get the same desired results.