Novel techniques and an efficient algorithm for closed pattern mining

作者：

Highlights：

• Frequent closed itemset mining and biclustering can be reduced to the same problem.

• A new and efficient algorithm for mining frequent closed patterns is presented.

• We introduce a unique approach to transform {-1,0,1}-type data into binary format.

• We propose an original aggregation method to detect the most meaningful patterns.

• We offer a novel technique for visualization of biclustering results.

摘要

•Frequent closed itemset mining and biclustering can be reduced to the same problem.•A new and efficient algorithm for mining frequent closed patterns is presented.•We introduce a unique approach to transform {-1,0,1}-type data into binary format.•We propose an original aggregation method to detect the most meaningful patterns.•We offer a novel technique for visualization of biclustering results.

论文关键词：Biclustering,Closed frequent itemset mining,Clustering visualization,Data mining algorithm,Pattern detection

论文评审过程：Available online 3 March 2014.

论文官网地址：https://doi.org/10.1016/j.eswa.2014.02.029