Unlock instant, AI-driven research and patent intelligence for your innovation.

A multi-label active learning classification method and system based on SVM

A technology of active learning and classification method, applied in the field of machine learning, it can solve problems such as inapplicability of multi-label samples, and achieve the effect of solving learning classification problems and saving manpower

Active Publication Date: 2017-06-16
SUZHOU UNIV
View PDF2 Cites 1 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0004] In view of this, this application provides a multi-label active learning classification method and system based on SVM, which is used to solve the problem that the existing active learning classification methods are not suitable for multi-label samples

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
  • A multi-label active learning classification method and system based on SVM
  • A multi-label active learning classification method and system based on SVM
  • A multi-label active learning classification method and system based on SVM

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0061] Embodiment 1 of the present invention discloses a multi-label active learning classification method based on SVM, see figure 1 as shown, figure 1 It is a flowchart of an SVM-based multi-label active learning classification method disclosed in Embodiment 1 of the present invention. The method includes:

[0062] S101: Construct a candidate sample set.

[0063] In this step, the candidate sample set is specifically a set of samples selected based on the Max-Margin uncertainty sampling strategy, where Max-Margin uncertainty is a sampling strategy based on uncertainty, and the uncertainty sampling strategy is to use The trained classifier classifies the samples, and selects those samples with high uncertainty through a certain selection criterion.

[0064] Such as figure 2 as shown, figure 2 It is a flow chart of constructing a candidate sample set disclosed in Embodiment 1 of the present invention. include:

[0065] S201: Train part of the training samples to obtai...

Embodiment 2

[0112] Embodiment 2 of the present invention discloses a multi-label active learning classification system based on SVM, see Figure 4 as shown, Figure 4 It is a schematic structural diagram of an SVM-based multi-label active learning classification system disclosed in Embodiment 2 of the present invention. The system includes: a construction unit 401, a determination unit 402, a labeling unit 403, an update unit 404 and a classification unit 405, wherein:

[0113] A construction unit 401, configured to construct a candidate sample set.

[0114] It should be noted that the construction unit 401 specifically uses the samples selected based on the Max-Margin uncertainty sampling strategy to construct the candidate sample set. Among them, Max-Margin uncertainty is a sampling strategy based on uncertainty. The uncertainty sampling strategy is to use the trained classifier to classify samples, and select those with high uncertainty through a certain selection standard. sample. ...

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 present application discloses an SVM-based multi-label active learning classification method and system, the method comprising: constructing a candidate sample set; determining the label set to which each sample in the candidate sample set belongs; labeling the label set to which each sample belongs in the sample set and the candidate sample set, and obtaining the labeled sample; adding the labeled sample to the training sample set for training, and updating the classifier; using the classifier to classify the acquired samples Classification. The SVM-based multi-label active learning classification method preliminarily determines the label set to which the sample belongs, thereby saving labor costs and manual labeling time to a large extent, thereby solving the multi-label problem on the basis of saving manpower. Sample learning classification problem.

Description

technical field [0001] The present application relates to the technical field of machine learning, and more specifically, relates to an SVM-based multi-label active learning classification method and system. Background technique [0002] With the advent of the information age, a large amount of information has begun to exist in a computer-readable form, and the amount has increased dramatically. However, the information is mixed, and many meaningful data are submerged by a large amount of junk information. How to automatically classify useful information from these information will be an important topic. [0003] In the existing classification problems, it is assumed that a sample belongs to only one class label, and then the sample is classified by using a support vector machine through a supervised learning method. Support Vector Machine (SVM, Support Vector Machine) is a statistical learning method that has been widely used since the 1990s. It is a new classification tec...

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 Patents(China)
IPC IPC(8): G06F17/30G06K9/62
CPCG06N20/00G06F18/2411
Inventor 赵朋朋焦阳鲜学丰吴健崔志明
Owner SUZHOU UNIV