Non-Linear Correlation Discovery-Based Technique in Data Mining

摘要

In this paper, a novel technique was present for mining complete correlation itemset, named as NLCD(non-linear correlation discovery) from hereon. In the first, it employ the vertical representation of a database, and then to find the direction between the correlative itemsets with fast processing and lots of them through the whole, including many kinds of correlation. Transaction ids of each itemset are mapped and compressed to discrete bool sequence. Lastly it was evaluated the algorithm against algorithms LCD using a variety of data sets with short and long frequent patterns. Experimental report showed that the NLCD algorithm outperforms.