Supercharge Your Innovation With Domain-Expert AI Agents!

New category recognition method based on deep learning

A recognition method and category technology, applied in the field of machine learning, can solve the problem of low accuracy in the abnormal detection stage, and achieve the effect of ensuring recall and precision, good general applicability, and improved recognition accuracy.

Active Publication Date: 2021-07-20
INST OF COMPUTING TECH CHINESE ACAD OF SCI
View PDF0 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

The main disadvantage of this method lies in the low accuracy of the anomaly detection stage, especially when dealing with data with temporal and spatial correlations, such as images, speech, text, natural language processing, 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
  • New category recognition method based on deep learning
  • New category recognition method based on deep learning
  • New category recognition method based on deep learning

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0026] In order to make the purpose, technical solution and advantages of the present invention more clear, the method for identifying new categories based on deep learning provided in the embodiments of the present invention will be described below with reference to the accompanying drawings.

[0027] Generally speaking, machine learning (including deep learning) can be divided into three types: supervised learning, unsupervised learning and semi-supervised learning. Among them, supervised learning means that the categories of all training samples are marked. The purpose of machine learning is to determine the category of predicted samples, and the result must belong to the existing category; unsupervised learning means that the categories of all training samples are unlabeled. , the purpose of unsupervised learning is to cluster the training samples, and further determine the categories of the predicted samples and which categories of training samples are most similar; the ca...

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 new category identification method based on deep learning, comprising the following steps: inputting the predicted samples into the deep network to obtain the output vector of the predicted samples; calculating the output vector of the predicted samples, and the training samples of each category represent the similarity value between the vectors, thereby identifying the category of the predicted sample; wherein, the representative vector of each category in the training sample is obtained by inputting the training sample set into the deep network to obtain the output vector set, and according to The output vector set of the training samples is calculated.

Description

technical field [0001] The invention relates to the technical field of machine learning, in particular to a new category recognition method based on deep learning. Background technique [0002] As a new type of machine learning method, deep learning has a good recognition effect on sample data associated with time and space domains, such as images, audio, text, etc., and is robust to transformations such as translation and deformation of sample data. Therefore, it has been widely used as soon as it was proposed. [0003] Although deep learning has obvious advantages in recognition accuracy, the complexity of the training phase is higher than that of classic machine learning methods such as SVM. Therefore, how to learn the optimal network model for a specific problem and make the network model suitable for a universal network structure is a difficult problem. For example, the prediction samples and training samples of traditional machine learning usually satisfy the strict ...

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): G06K9/62G06N3/08
CPCG06N3/08G06F18/23G06F18/22G06F18/24G06F18/214
Inventor 邢云冰陈益强蒋鑫龙
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