Integrated TSK fuzzy classifier based on IFCM, KNN and data dictionary

A data dictionary, fuzzy classification technology, applied in character and pattern recognition, instruments, computer parts, etc., can solve the problem of unreliable classifiers, verification data is not directly and fully utilized, and intermediate sub-classifiers have no interpretability and other issues to achieve the effect of easy addition or deletion

Active Publication Date: 2019-10-01
HUZHOU TEACHERS COLLEGE
View PDF3 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

In other words, directly training a TSK fuzzy classifier with raw training data or integrating a TSK fuzzy classifier (i.e., multiple subsets of raw training data) will obviously not be suitable for the scenarios considered above, due to the following reasons: (1) due to the noise The original training data, the classifier trained may be unreliable; (2) Most of the existing TSK fuzzy classifier training algorithms use the cross-validation strategy on the verification data to determine the relevant parameters, and the precious verification data has not been directly and Fully utilized; (3) When the traditional integrated TSK fuzzy classifier is used to classify the original data, due to the existence of intermediate variables, the intermediate sub-classifiers have no interpretability

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 more

Image

Smart Image Click on the blue labels to locate them in the text.
Viewing Examples
Smart Image
  • Integrated TSK fuzzy classifier based on IFCM, KNN and data dictionary
  • Integrated TSK fuzzy classifier based on IFCM, KNN and data dictionary
  • Integrated TSK fuzzy classifier based on IFCM, KNN and data dictionary

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0034] refer to figure 1 , a kind of integrated TSK fuzzy classifier based on IFCM, KNN and data dictionary of the present invention, it is characterized in that: comprise the following steps successively:

[0035] a) For each sub-classifier from D tr Randomly sample training data subset D in 1 ,D 2 ,...,D L , and D 1 ∪D 2 ∪……∪D L =D tr ;

[0036] b) Parallel training of L zero-order TSK fuzzy sub-classifiers; adopting the idea of ​​parallel learning to simultaneously train L zero-order TSK fuzzy sub-classifiers, specifically including the following steps:

[0037] b1. Assign the number of fuzzy rules to each sub-classifier;

[0038] b2. Construct 5 Gaussian membership functions, whose membership central point set is {0, 0.25, 0.5, 0.75, 1}, randomly assign a value from the central point set in each dimension and construct a regular combination matrix Construct the kernel width matrix by assigning each element a random positive number Use Gaussian membership funct...

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

No PUM Login to View More

Abstract

The invention discloses an integrated TSK fuzzy classifier based on IFCM, KNN and a data dictionary. A method sequentially comprises the following steps: randomly sampling a training data subset fromDtr for each sub-classifier; training l zero-order TSK fuzzy sub-classifiers in parallel; for each sample of the verification data set, calculating the value of each corresponding output function, taking the value as an enhancement feature, and combining the original feature with the enhancement feature to form an enhancement verification data set; after an IFCM algorithm is used for up-regulationof the enhanced verification data set, generating a representative center point and a corresponding tag thereof, and removing enhanced characteristics to obtain a data dictionary; and for any test sample, finding k nearest points on the data dictionary by using a KNN method, and determining the class label of the test sample based on a voting strategy. The classification method has high classification performance, high robustness and high interpretability.

Description

【Technical field】 [0001] The invention relates to the technical field of fuzzy recognition and machine learning, in particular to the technical field of integrated TSK rapid classification based on IFCM, KNN and data dictionary. 【Background technique】 [0002] Due to the powerful approximation ability and high interpretability of Takagi-Sugeno-Kang (TSK) fuzzy classifier, it has been successfully applied in many fields, such as pattern recognition, intelligent data processing and adaptive fuzzy control. However, in some scenarios, such as crowdsourcing problems, network image identification, etc., we usually have precious but small amounts of real data and their labels (referred to as validation data in this paper), and it is easy to collect a large number of original or even multi-source training Data, usually these training data inevitably contain noise data. But if a classification model is built on these training data, its performance will be severely degraded. In othe...

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
no application Login to View More
IPC IPC(8): G06K9/62
CPCG06F18/24G06F18/214
Inventor蒋云良张雄涛梁荣华王瑞琴
OwnerHUZHOU TEACHERS COLLEGE