Supercharge Your Innovation With Domain-Expert AI Agents!

Support vector machine fast realization method and device based on data compression expression

A technology of support vector machines and support vectors, which is applied to computer components, character and pattern recognition, instruments, etc., can solve the problem of large storage space occupied by support vectors, and achieve the effect of saving storage space and memory space

Active Publication Date: 2017-06-13
PEKING UNIV SHENZHEN GRADUATE SCHOOL
View PDF5 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0012] This application provides a method and device for quickly implementing a support vector machine based on data compression representation, which solves the problem that the support vector occupies a large storage space in the learning algorithm based on the support vector machine

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
  • Support vector machine fast realization method and device based on data compression expression
  • Support vector machine fast realization method and device based on data compression expression
  • Support vector machine fast realization method and device based on data compression expression

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0064] Please refer to figure 1 , the present embodiment provides a method for quickly implementing a support vector machine based on data compression representation, comprising the following steps:

[0065] S1.1: Make training samples and convert the data to be trained into a prescribed format.

[0066] S1.2: Train the training samples to obtain model data.

[0067] S1.3: Store each support vector component in the model data in memory with a size of 24 bits.

[0068] S1.4: Predict the category of the recognition target according to the model data, and obtain a prediction result.

[0069] In step S1.4, the model data should be imported into memory first before recognition, that is, step S1.3. Recognition refers to judging the category of a specific recognition target based on the existing model data. The model data contains important support vector data, and the following will describe how the method provided by this embodiment reduces the space occupied by the support vec...

Embodiment 2

[0086] Please refer to Figure 5 , based on the method for quickly implementing a support vector machine based on data compression representation provided in the first embodiment above, this embodiment provides a device for quickly implementing a support vector machine based on data compression representation, including a sample production module 101, a training module 102, Preservation module 103 and prediction module 104 .

[0087] The sample making module 101 is used to make training samples, and convert the data to be trained into a specified format.

[0088] The training module 102 is used for training the training samples to obtain model data.

[0089] The saving module 103 is used to save each support vector component in the model data in the memory with a size of 24 bits.

[0090] The prediction module 104 is used to predict the category of the recognition target according to the model data, and obtain a prediction result.

[0091] The prediction module 104 should fir...

Embodiment 3

[0106] Also refer to figure 1 , this implementation provides another fast implementation method of support vector machine based on data compression representation, including the following steps:

[0107] S1.1: Make training samples and convert the data to be trained into a prescribed format.

[0108] S1.2: Train the training samples to obtain model data.

[0109] S1.3: Store each support vector component in the model data in memory with a size of 24 bits.

[0110] S1.4: Predict the category of the recognition target according to the model data, and obtain a prediction result.

[0111] Firstly, the method provided in this embodiment is compared with the above-mentioned first embodiment, and step S1.3 is the same as that of the first embodiment, and will not be repeated in this embodiment. Therefore, since the support vector storage method provided by the first embodiment is adopted, this embodiment can also solve the problem of large support vector storage space in the prior...

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 provides a support vector machine fast realization method and device based on data compression expression, is suitable for circuit design of all logic IC or ASIC chips requiring saving of support vector storage space, including but not limited to image recognition and other application fields, and belongs to the technical field of circuit design of logic IC or ASIC chips. According to the method, when model data obtained after training is performed on a training sample is saved, each support vector component in the model data is saved in a memory at the size of 24 bits. Compared with the prior art, under the condition that the precision requirement of a support vector machine is met, each support vector component saves 8-bit memory space , and support vectors can be stored only in three quarters of an original storage area. Due to the fact that the number of the support vectors is generally large and the dimensionality of the support vectors is high, numerous storage space can be saved by the adoption of the support vector machine fast realization method based on data compression expression.

Description

technical field [0001] The present application relates to the field of computer data processing, in particular to a method and device for quickly implementing a support vector machine based on data compression representation. Background technique [0002] With the rapid development of the information industry and the popular awareness of public security, the field of video surveillance has attracted more and more attention from the public. The result of the increase of surveillance cameras is that higher requirements are placed on the data processing and storage capabilities of the system background. At the same time, it must consume a lot of manpower to manually identify specific targets in response to emergencies. If some identification processes are embedded in the foreground, the system's requirements for background performance will be greatly reduced. [0003] In recent years, face recognition, gesture recognition and other technologies have begun to be applied on a la...

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
CPCG06F18/2411G06F18/214
Inventor 袁誉乐赵勇贺思颖王新安
Owner PEKING UNIV SHENZHEN GRADUATE SCHOOL
Features
  • R&D
  • Intellectual Property
  • Life Sciences
  • Materials
  • Tech Scout
Why Patsnap Eureka
  • Unparalleled Data Quality
  • Higher Quality Content
  • 60% Fewer Hallucinations
Social media
Patsnap Eureka Blog
Learn More

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

© 2025 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap|About US| Contact US: help@patsnap.com