Eureka AIR delivers breakthrough ideas for toughest innovation challenges, trusted by R&D personnel around the world.

A feature screening method and device

A feature screening and feature set technology, applied in the field of data processing, can solve problems such as inability to apply dynamic changes of feature data sets

Active Publication Date: 2020-10-13
BEIJING INST OF ENVIRONMENTAL FEATURES
View PDF5 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

Due to various reasons such as data measurement errors and sensor condition limitations, each feature detected by the radar contains missing values, that is, the features detected by the radar each time change dynamically, but the existing feature selection models and algorithms for screening features It is designed for static data and cannot be applied to dynamic changes in feature data sets

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 feature screening method and device
  • A feature screening method and device
  • A feature screening method and device

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0102] Such as figure 1 As shown, the feature screening method provided by the embodiment of the present invention may include the following steps:

[0103] Step 101: Obtain at least two features to be screened;

[0104] Step 102: respectively determine the evaluation metric value corresponding to each feature;

[0105] Step 103: According to the correlation between each feature, divide at least two features into at least two feature subsets, wherein each feature subset includes at least one feature, and the corresponding evaluation metric value in each feature subset is the largest The correlation coefficient between one feature and other features is greater than the preset first correlation coefficient threshold;

[0106] Step 104: Obtain a feature with the largest corresponding evaluation metric value from each feature subset, and construct each acquired feature into an optimal feature set;

[0107] Step 105: Update the features included in the optimal feature set by cal...

Embodiment 2

[0111] On the basis of the feature screening method provided in Embodiment 1, the process of determining the evaluation metric value corresponding to each feature in step 102 can be specifically implemented through the following steps:

[0112] A1: Construct the first feature set including the acquired features, and obtain the classification set of the first feature set in the pre-obtained decision feature subset, where the decision feature subset is the sum of the pre-obtained decision feature set Subset;

[0113] Substitute the decision feature subset, classification set, and decision feature set into the following equation set 1 to obtain the first evaluation metric parameters corresponding to each feature, where equation set 1 includes:

[0114]

[0115] Among them, U represents the first feature set, and U={x 1 ,x 2 ,x 3 …x n}, n represents the total number of features; B represents the subset of decision-making features, D|B represents the classification set, and ...

Embodiment 3

[0124] On the basis of the feature screening method provided in Embodiment 2, the process of dividing each feature into at least two feature subsets in step 103, such as figure 2 As shown, it can be realized through the following steps:

[0125] Step 201: Substituting the samples included in each sample set into Equation 3 to obtain a set of correlation parameters of each feature relative to other features;

[0126] Equation set three includes:

[0127]

[0128] in, Characterizes the correlation parameter of the i-th feature with respect to the j-th feature; X j Represent a set consisting of samples corresponding to the jth feature in each sample set; F i Represent a set consisting of samples corresponding to the i-th feature in each sample set; W i Characterize the correlation parameter set corresponding to the i-th feature;

[0129] Step 202: For each feature pair including two features, calculate the mean value of the correlation parameter of the first feature rel...

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 the technical field of data processing, and provides a feature screening method and device. The method includes: acquiring at least two features to be screened; respectively determining the evaluation measurement value corresponding to each feature; , divide at least two features into at least two feature subsets, wherein, in each feature subset, the correlation coefficient between a feature with the largest evaluation metric value and other features is greater than the preset first correlation Coefficient threshold; obtain a feature with the largest corresponding evaluation value from each feature subset, and construct each acquired feature as an optimal feature set; by calculating the recognition probability of the optimal feature set, the features included in the optimal feature set performing an update; using each feature included in the optimal feature set after the update as a result of screening at least two features. This scheme can perform feature screening in the case of dynamic changes in feature data sets.

Description

technical field [0001] The invention relates to the technical field of data processing, in particular to a feature screening method and device. Background technique [0002] When identifying a space target through radar, the radar will detect multiple features of the space target, and then select each detected feature through a feature selection model and algorithm, and then use the selected features to identify the space target . Due to various reasons such as data measurement errors and sensor condition limitations, each feature detected by the radar contains missing values, that is, the features detected by the radar each time change dynamically, but the existing feature selection models and algorithms for screening features It is designed for static data and cannot be applied to situations where feature datasets change dynamically. [0003] Therefore, to address the above deficiencies, it is necessary to provide a feature screening scheme that can be applied to the dyn...

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
CPCG06F18/2115
Inventor 盛晶任红梅袁莉
Owner BEIJING INST OF ENVIRONMENTAL FEATURES
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
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