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

Logarithm normalization method for improving local prediction ability of neural network

A neural network and normalization technology, applied in neural learning methods, biological neural network models, electrical digital data processing, etc., can solve problems such as variable thresholds that are not considered, and achieve the effect of improving the success rate of prediction

Inactive Publication Date: 2016-05-04
TIANJIN UNIV
View PDF2 Cites 1 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0008] This method only considers the variation range of the variable, and does not consider the threshold that may exist in the variable

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
  • Logarithm normalization method for improving local prediction ability of neural network
  • Logarithm normalization method for improving local prediction ability of neural network
  • Logarithm normalization method for improving local prediction ability of neural network

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0020] The technical solution of the present invention will be described in further detail below in conjunction with the accompanying drawings and specific embodiments.

[0021] Such as image 3 Shown is a schematic diagram of a cockpit model of a specific embodiment of the present invention. The input of the neural network is the entrance velocity V of the cockpit model in , inlet temperature T in , entrance angle A in , entrance position L in and the exit location L out . For ease of calculation, the size of the entrance and exit is the same, and the optional positions of the entrance and exit are determined according to the size of the geometric model of the cockpit. There are 14 optional entrances and exits on the upper side of the window, and 16 optional entrances and exits on the lower side. 1 is the variation range and variation interval of each input variable.

[0022] Table 1. Neural network input variable parameters

[0023]

[0024] Neural network output ...

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 logarithm normalization method for improving local prediction ability of a neural network. The method comprises the following steps of: determining input and output of the neural network according to a prediction object, specifically, generating an input value of a training sample as a neural network input value through a sampling method, and calculating an output value of each training sample through a computational fluid mechanics method, namely, designing a target value, as a neural network output value; performing normalization for the neural network input value through a linear normalization method; determining a threshold value of the neural network output value, and performing normalization for the neural network output value through a logarithm normalization method. Compared with the conventional linear normalization method, the logarithm normalization method provided by the invention improves prediction success rate of the neural network by 17.1%.

Description

technical field [0001] The invention relates to the application field of artificial intelligence, in particular to a logarithmic normalization method for improving the local prediction ability of a neural network. Background technique [0002] Artificial neural network is proposed and developed on the basis of modern neuroscience, which can reflect a certain mathematical model of human brain structure and function. There are many structures of artificial neural networks, which can be mainly divided into feedforward neural networks, self-organizing competitive neural networks, and feedback neural networks. In the feed-forward neural network, the transmission of information is carried out layer by layer, that is, from the input layer to each hidden layer, and finally to the output layer. The input of the next layer is only related to the previous output, and generally there is no feedback. loop, so it is called a feed-forward network. The neurons in the feedforward network c...

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): G06F17/50G06N3/08
CPCG06F30/13G06N3/086
Inventor 尤学一张天虎
Owner TIANJIN 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