.png){kind=icon}
.png){kind=icon}
Return
Generalized objects in the system with dispersed knowledge
.png){kind=icon}
PRE
AI
DOI：10.1016/j.eswa.2020.113773.png)
.png){kind=icon}
Original request for help.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
The inference process based on a set of local decision tables is considered in the article. Determining global decisions in such a situation is a complex and time-consuming task. The aim of the presented approach is to simplify this process by significantly reducing the size of local decision tables and, at the same time, maintaining the quality of decisions made. The research objective is the application of generalized objects with respect to the indiscernibility relation. When we use the generalized objects only the relevant and consistent knowledge remains in tables. In addition, the use of generalized objects for local tables results in a significant reduction in the number of objects in the tables. Such a change has a huge impact on the system's construction for dispersed data. The paper introduces a definition of the generalized objects that are suitable for quantitative data. In addition, a definition of the generalized objects generated with the accuracy expressed by the parameter, which are appropriate for qualitative data, is given. It was experimentally tested that the use of generalized objects significantly reduces the number of objects in local tables, up to 80% or even 90% percent of the original table. In addition, it was shown that the use of generalized objects for local tables with large number of attributes gives comparable quality of classification to the dispersed system using full objects. Moreover, it was depicted that the quality of classification obtained for such data and the dispersed system with the generalized objects is better than the quality that are obtained for full objects without using the dispersed system. (C) 2020 Elsevier Ltd. All rights reserved.
Keywords:
Cooperative systems
Dispersed knowledge
Generalized objects
lndiscernibility relation
Knowledge based systems
AI Summary
Key information extracted from the uploaded paper, including a brief overview, abstract, background, key highlights, visual analysis, and future outlook.
Journal
IF:
7.5
Papers:
2.9W
Citations:
10.2W