Dirichlet distribution-based confidence degree calculation method

A technology of Dirichlet and calculation method, which is applied in the field of confidence calculation, can solve the problems of calculation result deviation, missing completion, inability to deal with confidence calculation problems, etc., and achieve the effect of reducing prediction

Inactive Publication Date: 2017-09-08
四川省公安科研中心 +4
View PDF4 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0002] The traditional regression confidence calculation method cannot deal with the confidence calculation problem in the case of high dimensionality and linear inseparability
As the number of features increases, it is difficult for traditional confidence calculation methods to perform missing completion and deviation correction for each feature probability, which may easily cause deviations in calculation results.

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
  • Dirichlet distribution-based confidence degree calculation method

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0027] The preferred embodiments of the present invention are given below in conjunction with the accompanying drawings to describe the technical solution of the present invention in detail.

[0028] Such as figure 1 As shown, the confidence calculation method based on Dirichlet distribution of the present invention comprises the following steps:

[0029] Step 1, hidden Dirichlet distribution parameter estimation training, used to obtain sample data and perform hidden Dirichlet distribution training to obtain model parameter estimates;

[0030] Step 2, input feature data, feature processing, used to calculate the predicted data probability, and correct the feature probability;

[0031] Step 3, parameter and weight normalization, confidence calculation, which is used to perform complex algorithm calculations on the forecast data to calculate confidence and perform normalization processing;

[0032] Step 4, calculating and sorting by confidence.

[0033] Described step one co...

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 Dirichlet distribution-based confidence degree calculation method. The method comprises the following steps of 1, estimating and training implicit Dirichlet distribution parameters; 2, inputting feature data and performing feature processing; 3, performing parameter and weight normalization and confidence degree calculation; and 4, performing calculation and sorting by utilizing confidence degree. According to the method, the problem of separation surface prediction under the condition that high dimension and linearity are non-separable can be solved; and the feature probability under the condition of deficiency, strong correlation and mutual exclusion is optimized.

Description

technical field [0001] The present invention relates to a confidence degree calculation method, in particular to a confidence degree calculation method based on Dirichlet distribution. Background technique [0002] The traditional regression confidence calculation method cannot deal with the confidence calculation problem in the case of high dimensionality and linear inseparability. As the number of features increases, it is difficult for traditional confidence calculation methods to perform missing completion and deviation correction for each feature probability, which may easily cause deviations in calculation results. Contents of the invention [0003] The technical problem to be solved by the present invention is to provide a confidence degree calculation method based on Dirichlet distribution, which can solve the interface prediction problem in the case of high-dimensional and linear inseparability, and optimize the interface prediction problem in the case of missing,...

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): G06F17/10
CPCG06F17/10
Inventor 李建闵圣捷贺晨阳杨伟华白云石葆梅张仕洪彭京赵敬千赖宇姜淮韬谢伯栋杨春勇周洋龙鑫郑镖
Owner 四川省公安科研中心
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