N.A. Korenevskiy, S.A. Filist, A.G. Ustinov, E.B. Ryabkova
The method presented in this work allows one to synthesize fuzzy decision rules for prediction and medical diagnostics. The suggested method is based on the fuzzy logic decision making.
The synthesis of the fuzzy decision rules is based on the classification of the information derived from the object. The classification is realized by exploratory data analysis. To solve this problem the separating hypersurfaces in the two-dimensional space are built. Taking into account these hypersurfaces we can decide to what class belongs the object. We can consider the distance between different structures of the investigated classes and between different points of the multidimensional space of attributes as basic variables. Taking into account this basic variables we can choose the class of the investigated object.
The results of the classification are estimated by the experts. To decrease the level of the subjectivism it’s necessary to solve the error minimization problem of the classification.