A parallel rough set attribute reduction algorithm based on attribute frequency

Chen Yanyun,Qiu Jianlin,Chen Jianping,Chen Li
DOI: https://doi.org/10.1109/FSKD.2012.6233881
2012-01-01
Abstract:Reduction is the core issue of rough set theory. Integrate the parallel idea into the attribute reduction in order to improve the efficiency and construct a parallel rough set attribute reduction algorithm based on attribute frequency. The algorithm divides the original information system into several subsystems according to the proposed division of thought, then uses the properties of frequency as the attribute importance to reduce each subsystem. When encountered with the situation of the same frequency, select the most uniform classification of the property by calculating the information entropy, and finally, obtain the global reduction from partial reduction. The algorithm is applied to corn breeding. Experiments show that the algorithm outperforms the traditional algorithms.
What problem does this paper attempt to address?