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

Neural network sample selection method and device based on active learning

A neural network and active learning technology, applied in the field of machine learning, can solve problems such as high algorithm complexity and inapplicability, achieve the effect of reducing the number of training samples, reducing time and cost, and improving classification effects

Inactive Publication Date: 2012-06-13
HOHAI UNIV
View PDF0 Cites 14 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

However, the complexity of this algorithm is extremely high, and it is not applicable to many complex problems.

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
  • Neural network sample selection method and device based on active learning
  • Neural network sample selection method and device based on active learning
  • Neural network sample selection method and device based on active learning

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0024] The present invention will be described in detail below in conjunction with the accompanying drawings and specific embodiments.

[0025] Taking the multi-layer perceptron neural network as an example, the method for selecting samples of the feedforward neural network according to the present invention is described. However, those skilled in the art should understand that the present invention is not limited to MLP neural networks, but can be applied to other feed-forward neural networks.

[0026] MLP is a fully connected feed-forward neural network suitable for object classification. The structure of the MLP is as figure 1 As shown, it is a three-layer feed-forward network: the input layer MA is composed of input pattern nodes, x i Represents the i-th component of the input pattern vector (i=1, 2, ..., n); the second layer is the hidden layer MB, which consists of m nodes b j (j=1, 2, . . . , m) composition. The third layer is the output layer MC, which consists of ...

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 a neural network sample selection method based on active learning, belonging to the field of machinery learning in intelligent science and technology. The neural network sample selection method comprises the following steps of: taking sensibility as a reference, and looking for sample points with severely-changed periphery through selecting the sample points with large sensibility, wherein these sample points are often important to a training classifier. The invention further discloses a neural network sample selection device based on active learning. According to the neural network sample selection method and device based on active learning, disclosed by the invention, the number of necessarily marked sample points can be effectively reduced, and the performance of a classifier is improved through reducing the marks.

Description

technical field [0001] The invention relates to a learning sample selection method and a device for neural network design, in particular to an active learning-based learning sample selection method and a device that can effectively improve neural network classification efficiency, and belongs to the field of machine learning in intelligent science and technology. Background technique [0002] When designing a neural network classifier, the labeling of training samples is generally done by experts, which often costs a lot of money and time. In the past, when selecting samples to be labeled, they were generally randomly selected from the obtained unlabeled samples. Training a classifier in this way often requires a large number of labeled samples, which requires a lot of manpower and material resources, and it is easy to prolong the training time. [0003] Active learning technology is to filter unlabeled samples through certain criteria, and then manually mark the obtained 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
no application Login to View More
IPC IPC(8): G06N3/08
Inventor 储荣王敏
Owner HOHAI UNIV
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