Supercharge Your Innovation With Domain-Expert AI Agents!

A method and system for transfer learning of the Internet of Things

A technology of transfer learning and Internet of Things, which is applied in the field of machine learning and transfer learning to achieve the effect of reducing the cost of manual labeling

Active Publication Date: 2021-04-27
INST OF COMPUTING TECH CHINESE ACAD OF SCI
View PDF4 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0005] Aiming at the problem that the above IoT migration learning method cannot make the learner (Learner) perceive the node to recognize the confusion mode of the teacher (Teacher) perceive the node

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 method and system for transfer learning of the Internet of Things
  • A method and system for transfer learning of the Internet of Things
  • A method and system for transfer learning of the Internet of Things

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0038] The core content of the present invention is to automatically find and discover the advantageous ability sample groups in the learner's characteristic domain relative to the professor's perception node in the scenario of the migration of the Internet of Things, and further divide them into several advantageous ability groups. The representative samples of each dominant ability group are marked in a semi-supervised manner, and semi-labeled modification is performed on all learner sample data sets, so that learners can specify their dominant ability samples, and then in their dominant ability. The sample label has a labeling accuracy that is higher than the professor's ability to identify. The learner pattern recognition model trained by this method can have the ability to distinguish the confusion pattern of the teacher's perception node.

[0039] The pattern recognition ability of the professor node is limited, and some patterns in the sample set cannot be correctly ide...

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 invention relates to a transfer learning method and system for the Internet of Things, including: when the professor and the learner perceive nodes are deployed at the same time, the professor assists the learner to mark the current sample by identifying and transmitting the semi-label in real time, and using the semi-label The properties of the labels find patterns and corresponding groups of samples that are distinguishable in the learner dataset but not by the teacher classifier. Using a semi-supervised method to seek artificial auxiliary labels, the system will use the half-label vector value of its initial state, the half-label value of surrounding samples, and the most similar typical samples with semi-supervised labels for each dominant ability sample in the learner dataset. The label value is modified by half-label, and the label value corresponding to the highest component in the sample half-label vector is assigned to each sample, so as to obtain the modified label sample set to train the learner pattern recognition model, so as to realize the pattern recognition ability from Transfer from teacher to learner.

Description

technical field [0001] The invention belongs to the field of machine learning and migration learning, and in particular relates to a method and system for migration learning of the Internet of Things. Background technique [0002] In the real-world IoT system, the types and number of IoT nodes in the existing structure often cannot meet the requirements of the corresponding scenarios and system objectives, so it is necessary to add new sensing nodes to enrich the sensing capabilities of the IoT system. When a new sensing node is added, the system often needs a large amount of labeled training and test data to build a machine learning model for the newly added node, but labeling a large amount of training and test data is a time-consuming and manual labeling cost. work. Therefore, it is a very meaningful work to transfer the pattern recognition capability of the existing sensing nodes to the newly added sensing nodes. The mode can be, for example: 1. Motion recognition mode...

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): G06K9/62
CPCG06F18/24137G06F18/24133G06F18/214
Inventor 王念崔莉赵泽
Owner INST OF COMPUTING TECH CHINESE ACAD OF SCI
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