Amaury Caballero, Kang Yen, Jose L. Abreu, Aldo Pardo
Method for Optimizing the Number and Precision of Interval-Valued Parameters in a Multi-Object System
Information Systems, Classification, Databases, Diffuse Information
In any decision-making process, it is necessary to evaluate the merit of different parameters and disclose the effect they have on the solution in order to optimize it. If the criteria are mathematically quantifiable, a mathematical model may be created for the evaluation process. The application of rough sets, neural networks, fuzzy logic or information theory are widely used mathematical tools for the solution of this type of problem. In this work two tasks are approached: (1) To minimize the number of required parameters(attributes) by discriminating different objects (classes), for the case when there is an overlap in the collected information from parameters (interval-valued information); (2) To calculate the minimum accuracy necessary in the selected attributes to discriminate all of the objects. This approach is very useful in both reducing the cost of the communication channels and in eliminating unnecessary stored information.
Cite this paper
Amaury Caballero, Kang Yen, Jose L. Abreu, Aldo Pardo. (2016) Method for Optimizing the Number and Precision of Interval-Valued Parameters in a Multi-Object System. International Journal of Control Systems and Robotics, 1, 139-143