Looking for breakthrough ideas for innovation challenges? Try Patsnap Eureka!

Multi-mark classification method, device, medium and computing device

A classification method and multi-label technology, applied in computing, computer components, instruments, etc., can solve problems such as inaccurate classification results, and achieve the effect of enriching data and semantics

Inactive Publication Date: 2017-11-03
XIAMEN UNIV OF TECH
View PDF0 Cites 6 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0005] The embodiment of the present application provides a multi-label classification method, device, medium, and computing equipment to solve the problem in the prior art that the multi-label problem is simply regarded as a combination of multiple single-label problems for multi-label classification, resulting in classification results inaccuracies etc.

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
  • Multi-mark classification method, device, medium and computing device
  • Multi-mark classification method, device, medium and computing device
  • Multi-mark classification method, device, medium and computing device

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0029] refer to figure 1 , which is a schematic flow chart of the multi-label classification method provided in Embodiment 1 of the present application, the method includes the following steps:

[0030] Step 101: For each label in the label set, determine the original positive example set and the original negative example set of the label; wherein, for each sample, if the sample has the label, the sample belongs to the original positive example of the label set, otherwise, the sample belongs to the labeled original set of negative examples.

[0031] Step 102: Carry out class alignment on the original positive example set and the original negative example set of each mark respectively, and obtain the positive example set after the class alignment of each mark and the negative example set after the class alignment; wherein, the class-aligned set of each mark The number of samples in the positive example set is equal, and the number of samples in the negative example set after t...

Embodiment 2

[0084] As shown in Table 1, it is a multi-label data set. There are a total of 6 samples in this table, and are x 1 ,x 2 ,...,x 6 , the mark set is {l 1 , l 2}.

[0085] Table 1 Samples and the marks they have

[0086]

[0087] Step 1: Class Alignment:

[0088] It can be seen from the statistical table 1 that l 1 The original set of positive examples is Its original set of negative examples is l 2 The original set of positive examples is Its original set of negative examples is

[0089] Obviously, (i.e. l 1 The number of positive samples of is 2), therefore In order to achieve the and For class alignment, you only need to add a positive example to Just go there. optional l 1 2 positive examples in x 2 and x 3 to generate positive examples (x 2 +x 3 ) / 2. Similarly, in this example l 2There are fewer negative examples, and class alignment is also required. You can choose l 2 2 negative examples of x 3 and x 4 to generate negative examp...

Embodiment 3

[0113] Based on the same inventive concept, the embodiment of the present application also provides a multi-label classification device, such as figure 2 Shown is a schematic structural diagram of the device, including:

[0114] The positive and negative example set determination module 201 is used to determine the original positive example set and the original negative example set of the label for each label in the label set; wherein, for each sample, if the sample has the label, the The sample belongs to the original positive example set of the label, otherwise, the sample belongs to the original negative example set of the label;

[0115] The class alignment module 202 is used to perform class alignment on the original positive example set and the original negative example set of each mark respectively, and obtain the positive example set after the class alignment of each mark and the negative example set after the class alignment; wherein, each marked The number of sampl...

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 belongs to the machine learning technical field and relates to a multi-mark classification method, a multi-mark classification device, a medium and a computing device. According to the embodiments of the present invention, after the original positive example set and original negative example set of marks are obtained; by means of classification alignment, specific attributes and operation for inserting specific attributes of related marks are determined, so that the specific attributes can be adopted to indicate correlations between the marks, and therefore, the data and semantics of the marks can be enriched; and therefore, multi-mark classification is more accurate than a method only adopting a single mark in the prior art. For example, deserts are correlated to camels, a picture which is dominated by a camel and contains a small part of a desert is classified as a desert picture; and if a picture contains a lake at dusk, and if the reflection of a sinking sun is in the lake, the picture would be classified as a lake picture in the prior art, while, the reflection of the sun in the lake is correlated to the sinking sun, the picture will be classified as a dusk scenery picture by the method of the invention.

Description

technical field [0001] The present application relates to the technical field of machine learning, in particular to a multi-label classification method, device, medium and computing equipment. Background technique [0002] The multi-label problem is widespread in machine learning. For example, in the image labeling problem, given labels such as "boat", "water", "mountain", "bridge", "pedestrian", "sunset", "cloud", a picture describing the scenery of the river can be Mark with one or more of these tags. For another example, in gene function classification, a gene may be related to markers such as "energy" and "metabolism" used to represent functional categories. Due to the large amount of marking work and the slow speed of manual marking, it is unrealistic to use manual marking. Therefore, it is particularly important to study the use of computer technology for automatic multi-label classification. [0003] In related technologies, an object that needs to be marked (abbr...

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
Patent Type & Authority Applications(China)
IPC IPC(8): G06K9/62
CPCG06F18/24137
Inventor 翁伟朱顺痣钟瑛李建敏
Owner XIAMEN UNIV OF TECH
Who we serve
  • R&D Engineer
  • R&D Manager
  • IP Professional
Why Patsnap Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Patsnap Eureka Blog
Learn More
PatSnap group products

Browse by: Latest US Patents, China's latest patents, Technical Efficacy Thesaurus, Application Domain, Technology Topic, Popular Technical Reports.

© 2024 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap|About US| Contact US: help@patsnap.com