MapReduce based improved quick reduct algorithm with granular refinement using vertical partitioning scheme
作者：
Highlights：
• MapReduce based attribute reduction algorithm is proposed using Rough Set Theory.
• Uses Vertical partitioning scheme, that divides input data over attribute space.
• Achieves huge reduction in data transfer of Shuffle and Sort phase of MapReduce.
• Introduces Granular refinement in MapReduce based reduct computation.
• Scalable for data sets having larger attribute space, useful in Bioinformatics.
摘要
•MapReduce based attribute reduction algorithm is proposed using Rough Set Theory.•Uses Vertical partitioning scheme, that divides input data over attribute space.•Achieves huge reduction in data transfer of Shuffle and Sort phase of MapReduce.•Introduces Granular refinement in MapReduce based reduct computation.•Scalable for data sets having larger attribute space, useful in Bioinformatics.
论文关键词：Rough sets,MapReduce,Apache spark,Reduct,Horizontal partitioning,Vertical partitioning,Feature subset selection
论文评审过程：Received 7 February 2019, Revised 7 October 2019, Accepted 9 October 2019, Available online 14 October 2019, Version of Record 16 January 2020.
论文官网地址：https://doi.org/10.1016/j.knosys.2019.105104
