Granular-Rule Extraction to Simplify Data

摘要

Granulation simplifies the data to better understand its complexity. It comforts this understanding by extracting the structure of data, essentially in big data or cloud computing scales. It can extract a simple granular-rules set from a complex data set. Granulation is associated with theory of fuzzy information granulation, which can be supported by fuzzy C-mean clustering. However, intersections of fuzzy clusters create redundant granular-rules. This paper proposes a granular-rules extraction method to simplify a data set into a granular-rule set with unique granular-rules. It performs based on two stages to construct and prune the granular-rules. We use four data sets to reveal the results, i.e., wine, servo, iris, and concrete compressive strength. The results reveal the ability of proposed method to simplify data sets by 58% to 91%.