Fuzzy pattern recognition method based on fuzzy n-cell number

Abstract

The invention discloses a fuzzy identification method based on a fuzzy ellipsoid book. The fuzzy identification method comprises the following steps: step 1, obtaining a to-be-identified target or observation data sets of l standard categories; Step 2, respectively calculating the mean value of each feature of the observation data set of the to-be-identified target and the standard category in thestep 1; and step 2, respectively calculating the mean value of each feature of the observation data set of the to-be-identified target and the standard category in the step 1. step 3, constructing afuzzy n-cell number in a standard category, and constructing fuzzy n-cell numbers of the to-be-identified target; step 4, respectively calculating centroids of the fuzzy n-cell number of the to-be-identified target and the fuzzy n-cell number in each standard category, calculating the Euclidean distance through the centroid of the to-be-identified target and the and the centroid of each standard category; step 5, according to the Euclidean distance value obtained in the step 4, judging the pattern classification of the to-be-identified target by using a minimum distance principle. According tothe invention, the fuzzy mode identification accuracy of the fuzzy object is improved.

Description

technical field [0001] The invention belongs to the field of fuzzy pattern recognition, in particular to a fuzzy pattern recognition method based on fuzzy n-cell numbers. Background technique [0002] Pattern recognition is a basic problem that is often encountered and dealt with in science, engineering, economy, society and even life. The mathematical form of this problem is to identify which standard class the target to be tested belongs to under the premise of artificially marking each standard class. At this time, it is necessary to formalize each object mathematically, and sometimes the formalization of the object has a fuzzy nature, and at this time, fuzzy mathematical methods must be used for recognition. [0003] For fuzzy pattern recognition with multi-factor uncertain information, triangular fuzzy numbers are often used to construct the membership function of multidimensional fuzzy numbers, and the measurement of pattern recognition is constructed around the conce...

AI Technical Summary

Problems solved by technology

[0004] In order to solve the problem that each component of the fuzzy number mean only depends on its corresponding fuzzy number component, and has nothing to do with other components, and the fuzzy number mean cannot reflect the overall structure of the multi-dimensional fuzzy number well, so that there is a certain deviation in the recognition results. The present invention provides a fuzzy pattern recognition method based on fuzzy n-cell numbers, which can be based on the centroid value of fuzzy n-cell numbers and the corresponding distance measure in an uncertain or imprecise environment, and then use the minimum distance principle to Multi-channel uncertain digital information for pattern recognition

Images