Group decision voting type membership function and pattern classification method based on group decision voting type membership function
A membership function and voting-type technology, which is applied in the field of group decision-making voting membership function and pattern classification based on it, can solve the problems of no membership function, difficult realization of equivalence division of upper approximation set, rough pattern classification, etc., to improve accuracy sexual effect
Pending Publication Date: 2020-08-25
SUQIAN COLLEGE
View PDF0 Cites 0 Cited by
- Summary
- Abstract
- Description
- Claims
- Application Information
AI Technical Summary
Problems solved by technology
[0002] At present, pattern classification is mostly carried out using parametric or non-parametric techniques such as statistical methods and stochastic processes. Data processing technologies such as artificial intelligence, machine learning, support vector machines, and neural networks are continuously introduced to make pattern classification capabilities more advanced in terms of classification granularity and accuracy. A big improvement has been obtained, but these methods are only suitable for large data samples, and seem powerless for small data samples, and the calculation is more complicated, and the speed cannot be guaranteed
[0003] Although the rapid pattern classification of small samples can be achieved with the help of traditional fuzzy set thinking, the membership function that determines the classification granularity and accuracy is often obtained through statistical methods or expert experience, and cannot adapt to pattern classification of different background conditions and different object characteristics. It cannot reflect the difference between the group consensus and individual understanding of the same object, and the rough set realizes the relatively fine division of information units through the equivalence class, but it is not easy to find the equivalence relationship, which makes the equivalence division of the upper approximation set very difficult. Difficult to achieve
[0004] Group decision-making and voting models better reflect people's common understanding and individual understanding of the same object. Through voting and decision-making, finer-grained information identification can be achieved, but the group decision-making voting model has not yet been applied to build a membership that reflects group consensus and individual cognition. function, and there is no group decision-making voting membership function to achieve rough approximation of the pattern classification of objects in the set
Method used
the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
View moreImage
Smart Image Click on the blue labels to locate them in the text.
Smart ImageViewing Examples
Examples
Experimental program
Comparison scheme
Effect test
Embodiment
[0038] Embodiment: take the color block color pattern recognition as an example to provide an exemplary description of the present invention, for a given RGB color block sample set F={color block 1→[204,0,1], color block 2→[153 ,0,0], color block 3→[255,102,102], color block 4→[51,0,0], color block 5→[255,153,153], color block 6→[102,0,0], color block 7→ [255,204,204], color block 8 → [255,0,0], color block 9 → [255,51,51]}, construct a group decision-making voting membership function and classify the color blocks based on this. Such as figure 1 -3, the embodiment of the present invention provides a group decision-making voting membership function construction method, including the following steps:
[0039] 1) Given a data object universe and a natural language descriptor set, apply any specified membership function to the data object universe to obtain a quasi-membership universe;
[0040] The 9 color blocks in the color block sample set F are all in RGB format, and the col...
the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More PUM
Login to View More Abstract
The invention discloses a group decision voting type membership function and a pattern classification method based on the group decision voting type membership function, and the method comprises the steps: applying any specified membership function to a data object discourse domain under a specified natural language descriptor, and obtaining a quasi-membership discourse domain; selecting an evaluation standard set to align to the membership degree to carry out voting evaluation decision; calculating the evaluation validity and the evaluation cardinal number of the evaluation standard; calculating a membership function value corresponding to the quasi membership degree; drawing a group decision voting type membership function image on the quasi-membership degree discourse domain; constructing a lower approximation set, an upper approximation set and an irrelevant set according to the voting evaluation result of the evaluation standard alignment membership degree, and respectively forming a consistent relevant category, a partial relevant category and an irrelevant category; combining the membership function image and a given horizontal threshold line to subdivide the quasi-membership discourse domain into a plurality of modes. According to the method, a membership function reflecting group consensus and individual understanding is provided, and the accuracy of pattern classification is improved.
Description
technical field [0001] The invention relates to a pattern classification method, in particular to a group decision-making voting membership function and a pattern classification method based on it, belonging to the technical field of pattern classification. Background technique [0002] At present, pattern classification is mostly carried out using parametric or non-parametric techniques such as statistical methods and stochastic processes. Data processing technologies such as artificial intelligence, machine learning, support vector machines, and neural networks are continuously introduced to make pattern classification capabilities more advanced in terms of classification granularity and accuracy. A large improvement has been obtained, but these methods are only suitable for large data samples, and seem powerless for small data samples, and the calculation is more complicated, and the speed cannot be guaranteed. [0003] Although the rapid pattern classification of small s...
Claims
the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More Application Information
Patent Timeline
Login to View More Patent Type & AuthorityApplications(China)
IPC IPC(8): G06K9/62
CPCG06F18/24
Inventor李守军李光宇邱义臻刘祎
OwnerSUQIAN COLLEGE