Two Different Views for Generalized Rough Sets with Applications

Rough set philosophy is a significant methodology in the knowledge discovery of databases. In the present paper, we suggest new sorts of rough set approximations using a multi-knowledge base; that is, a family of the finite number of general binary relations via different methods. The proposed methods depend basically on a new neighborhood (called basic-neighborhood). Generalized rough approximations (so-called, basic-approximations) represent a generalization to Pawlak's rough sets and some of their extensions as confirming in the present paper. We prove that the accuracy of the suggested approximations is the best. Many comparisons between these approaches and the previous methods are introduced. The main goal of the suggested techniques was to study the multi-information systems in order to extend the application field of rough set models. Thus, two important real-life applications are discussed to illustrate the importance of these methods. We applied the introduced approximations in a set-valued ordered information system in order to be accurate tools for decision-making. To illustrate our methods, we applied them to find the key foods that are healthy in nutrition modeling, as well as in the medical field to make a good decision regarding the heart attacks problem.

Automated summary

The paper introduces new rough set approximations using a multi-knowledge base, based on a new neighborhood called basic-neighborhood. These proposed methods are a generalization of Pawlak's rough sets and some of their extensions, proving their best accuracy. The main goal is to study multi-information systems and extend the application field of rough set models, with discussions on real-life applications like nutrition modeling and decision-making in the medical field.