INTELLIGENT CLASSIFICATION OF BIOPHYSICAL SYSTEM STATES USING FUZZY INTERVAL LOGIC
Iryna Tvoroshenko, Volodymyr Gorokhovatskyi · Telecommunications and Radio Engineering · 2019
The task of increasing the reliability of the adoption of classification decisions on the state of the biophysical system with fuzzy interval representations about the characteristics or properties of objects is solved. A formalized model for classifying the states of the research object is proposed. The model provides a mechanism for calculating the confidence coefficient for each situation from a defined set of space states of the system and provides an opportunity to present the investigated features of objects based on four types of membership functions, thus ensuring that the inaccuracies, fuzziness or unreliability of the available data and knowledge are eliminated. The proposed modification of the state classification method generates and evaluates several alternatives according to criteria when making classification decisions. Experimental testing of the system has been performed through software simulation, as well as costs, required for the software product development, have been calculated. The increase of the state classification reliability has been confirmed, the versatility of the proposed intellectual methods for the arbitrary set of fuzzy data has been established.