This is a Demo Server. Data inside this system is only for test purpose.

Scopus

Permanent URI for this collectionhttps://tuit-demo.gcris.com/handle/123456789/23

Browse

Browsing Scopus by Subject "classification"

Filter results by typing the first few letters
Now showing 1 - 1 of 1
  • Thumbnail Image
    Conference Object
    Citation Count: 0
    Classification Algorithms With a Complex Structure in Pattern Recognition
    (IEEE Computer Society, 2024) Nishanov, A.Kh.; Allamov, O.; Rakhmanov, A.T.; Yusupova, J.; Ruzibaev, O.B.; Department of System and Applied Programming
    The article considers in the intellectual processing of information, moving from a large unit of measurement to a smaller, more important unit of measurement, selecting sets of informative symbols and classifying symbols, developing modern recognition systems that help specialists in the field, as well as evaluating the states of objects to be classified, developing a classification algorithm for multi-class cases of educational samples. The development of a multi-class algorithm for symbol classification serves to develop the theory of symbol recognition. It deals with the construction of a decision rule construction algorithm. Its main purpose is to determine which of the given classes the recognized object belongs to. The complexity or simplicity of the proposed algorithm depends on the number of classes participating in the study sample. If the number of classes is two, then the formulation of decision rules is simple. That is why the classification algorithms proposed by most authors are solved for cases where the number of classes is three or more, the decisive rule in the matter of classification of objects is solved by making two classes. The construction of a multi-class decision rule for classification, which takes into account all classes at the same time, is a complex and urgent issue. In the article, in the classification of objects, a multi-level complex structured decision rule was developed based on special inequalities, and two-, three- and multi-class classification algorithms were developed. © 2024 IEEE.