An Effective Rough Neutrosophic Based Approach for Data Pre-Processing

Abstract

The core challenge in the process of knowledge discovery is the pre-processing of data. Pre-processing of data involves feature selection due to substantial amount of data and degree of dimension attribute. The concept of attribute reduction is introduced by Pawlak rough set theory for computational efficiency and accuracy. This process helps eliminate unnecessary attributes. Rough neutrosophic set is an extension of traditional rough set by hybrid of rough set and neutrosophic set theory. This paper introduced novel approach by integrating rough set and neutrosophic set to reduce data dependency. The formal principle is introduced for rough neutrosophic attribute reduction by using indiscernibility relation. Besides, we proposed the effective of rough neutrosophic set theory in selecting features. For the purpose of gaining a better understanding of proposed method, the algorithm of heuristic reduction is constructed for attribute reduction for data pre-processing. It contains seven steps of distributed jobs to produce the final output which is reduct. The result shows the impact of rough neutrosophic set on attributes, especially in term of dependency and reduct.