Neural network model reference neural cognitive mechanism and machine learning mathematical method

A neural network model and machine learning technology, applied in the field of machine learning, can solve problems such as data overfitting, achieve the effects of improving generalization, breaking through the limitations of big data, and improving scope and capabilities

Pending Publication Date: 2022-01-11
杭州翔毅科技有限公司
View PDF0 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0005] Aiming at the deficiencies of the prior art, the present invention provides a neural network model that draws on neurocognitive mechanisms and machine learning mathematical methods, and solves the problem that a large amount of data interconnected by DNN may be overfitting

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 model reference neural cognitive mechanism and machine learning mathematical method
  • Neural network model reference neural cognitive mechanism and machine learning mathematical method
  • Neural network model reference neural cognitive mechanism and machine learning mathematical method

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0038] like Figure 1-Figure 4 As shown, the embodiment of the present invention provides a neural network model for reference to neurocognitive mechanisms and machine learning mathematical methods, including the following steps:

[0039] S1: Establish and store the sample information in a sample library, generate adversarial samples through the fast gradient symbol method, divide the adversarial samples with different typical characteristics into different types, use the Kennard-Stone algorithm to classify the adversarial samples, and divide the adversarial samples into training sets and the test set;

[0040] S2: Enhance the migration ability of the adversarial sample library to the training set through lifelong continuous transfer learning;

[0041] S3: Connect the recursive neural network through the convolutional neural network, then connect the BP neural network and the naive Bayesian algorithm to establish a neural network model, and test the accuracy of the neural net...

Embodiment 2

[0059] like Figure 4 As shown, the neural network model refers to the system of neurocognitive mechanism and machine learning mathematical method, including an input module, an output module, a memory and a processor, the input module is electrically connected to the processor, and the storage module and the processor are electrically connected to each other. connected, the processor is electrically connected to the output module.

[0060] The storage module is electrically connected to the processor, the neural network model and sample library that can be run by the processor are stored in the memory, and the output module is used to output the data table.

[0061] When the processor is running, it executes the steps of S1--S4 in the method of this specification.

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 neural network model reference neural cognitive mechanism and machine learning mathematical method, and relates to the field of machine learning. The method comprises the following steps: S1, establishing a sample library for sample information, storing the sample library, generating adversarial samples through a fast gradient symbol method, and dividing the adversarial samples into a training set and a test set; S2, enhancing the migration ability from the adversarial sample library to the training set through lifelong continuous migration learning; S3, connecting a recurrent neural network through a convolutional neural network, and then connecting a BP neural network and a naive Bayes algorithm to establish a neural network model; and S4, classifying the target information in the adversarial sample library to form a data table. The adversarial samples are generated through the fast gradient symbol method, videos, images and texts are combined, real scene dialogues, dialogues with images and text expression are more specific, the artificial intelligence application range is widened, the observation thinking and expression ability is improved, and generalization is improved through different structures.

Description

technical field [0001] The invention relates to the field of machine learning, specifically, the neural network model uses neurocognitive mechanism and machine learning mathematical method for reference. Background technique [0002] Since the advent of the mathematical method of simulating the actual human neural network, people have gradually become accustomed to directly calling this artificial neural network a neural network. Neural network has broad and attractive prospects in the fields of system identification, pattern recognition, intelligent control, etc. Especially in intelligent control, people are particularly interested in the self-learning function of neural network, and regard this important feature of neural network as a It is one of the keys to solve the difficult problem of controller adaptability in automatic control. [0003] Deep neural network (DNN) is a multi-layer perceptron architecture used to solve complex learning problems. However, DNN faces cha...

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): G06N3/04G06N3/08G06V10/764G06V10/774
CPCG06N3/084G06N3/08G06N3/045G06F18/214G06F18/24155
Inventor 陈子轩雷铭轩郑正华国明李禅郭尚
Owner 杭州翔毅科技有限公司
Who we serve
  • R&D Engineer
  • R&D Manager
  • IP Professional
Why Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Try Eureka
PatSnap group products

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

© 2024 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap