资 源 简 介
Introduction
Dataset is a transactional database consisting of list of transactions and each of them containing a finite set of items. Dimensionality reduction is the process of finding a set of new items (factor-items) which is considerably smaller than the original set. These factor-items aims to comprise full or nearly full information about the original elements. This algorithm has been originally proposed by Petr Krajca, Jan Outrata, and Vilem Vychodil [1].
In [1], Krajca et al. show that frequent closed itemset can be used to efficiently find factor-items and thus accomplishing data dimensionality reduction. Also, if only frequent closed itemsets is not enough, they propose an on-demand solution to find additional factor-items until a configurable approximation degree is achieved.
To identify frequent closed itemsets this project also implements the LCM algorithm [2] origi