Preprocessing of Datasets Using Sequential and Parallel Approach: A Comparison

Abstract

Data preprocessing is a technique in data mining to make the data read for further processing according to the requirement. Preprocessing is required because the data might be incomplete, redundant, come from different sources which may require aggregation, etc., and data can be processed either sequentially or in parallel. There are several parallel frameworks such as Hadoop, MPI, and CUDA to process the data. A survey has been done to understand these parallel frameworks, and a comparison between sequential and parallel approach is carried out to compare the efficiency of the two approaches.