11-09-2017, 04:19 PM
One objective of the paper is to discuss a state-of-the-art methodology and fuzzy set algorithms in the field of pattern recognition. In problems of recognition and classification in the real world we face with blur that is related to various facets of human cognitive activity. An origin of the diffusion sources is related to the labels expressed in the space of characteristics, as well as with the labels of the classes that are taken into account in the procedures of classification. An obvious difference between a form of information processing is explained in detail by the theory of probabilities and fuzzy sets and a way of interpreting the results. In the sequel, pattern recognition methods are studied in two main streams, namely supervised and unsupervised learning. Different approaches to designing classification schemes (eg relationship calculation, decision-making approach, etc.) are taken into account. A method of selection of characteristics is introduced with the aid of diffuse and diffuse integral.