Unlock instant, AI-driven research and patent intelligence for your innovation.
Training sample screening method and device, electronic equipment and storage medium
What is Al technical title?
Al technical title is built by PatSnap Al team. It summarizes the technical point description of the patent document.
A technology for training samples and screening methods, applied in the Internet field
Pending Publication Date: 2020-11-03
BEIJING SANKUAI ONLINE TECH CO LTD
View PDF0 Cites 2 Cited by
Summary
Abstract
Description
Claims
Application Information
AI Technical Summary
This helps you quickly interpret patents by identifying the three key elements:
Problems solved by technology
Method used
Benefits of technology
Problems solved by technology
The time-consuming calculation of building a map of a large number of nodes has become the bottleneck of real-time deployment of online microservices
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
Click on the blue label to locate the original text in one second.
Reading with bidirectional positioning of images and text.
Smart Image
Examples
Experimental program
Comparison scheme
Effect test
Embodiment 1
[0090] refer to figure 1 , shows a flow chart of the steps of a training sample screening method provided by an embodiment of the present disclosure, as shown in figure 1 As shown, the training sample screening method may specifically include the following steps:
[0091] Step 101: Determine the sample set to be screened according to the correlation between the sample features of any two training samples in the training sample set.
[0094] There are multiple training samples included in the training sample set, wherein the multiple can be hundreds, thousands (for example, 1000, 2000, 3000, etc.), tens of thousands (for example, 10000, 20000, 30000, etc.), etc., Specifically, it can be determined according to b...
Embodiment 2
[0119] figure 2 It is a flow chart of the detailed steps of step 101. Step 101 may include: step 201 , step 202 , step 203 and step 204 .
[0120] Step 201: Construct a training sample graph according to each training sample in the training sample set; each of the training samples is a node on the training sample graph.
[0121] Embodiments of the present disclosure may be applied in a scenario of screening samples trained by a graph convolutional neural network.
[0122] The training sample set refers to the pre-acquired samples used to train the graph convolutional neural network.
[0123] The training sample set contains a plurality of training samples, wherein the number can be hundreds (for example, 500, 800, etc.), thousands (for example, 2000, 4000, etc.), tens of thousands (for example, 20000, 50000, etc. ), etc. Specifically, it may be determined according to business requirements.
[0124] After the training sample set is obtained, a training sample graph can be...
Embodiment 3
[0210] refer to Figure 8 , which shows a schematic structural diagram of a training sample screening device provided by an embodiment of the present disclosure, as shown in Figure 8 As shown, the training sample screening device 800 may include: a screening sample set determination module 810, a candidate sample set generation module 820, a label information entropy determination module 830, and a target sample screening module 840, wherein,
[0211] The screening sample set determination module 810 is configured to determine the sample set to be screened according to the correlation between the sample features of any two training samples in the training sample set.
[0212] Embodiments of the present disclosure may be applied in a scenario of screening samples trained by a graph convolutional neural network.
[0213] The training sample set refers to the pre-acquired samples used to train the graph convolutional neural network.
[0214] There are multiple training samples...
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
Login to View More
Abstract
The invention provides a training sample screening method and device, electronic equipment and a computer readable storage medium. The method comprises the steps of determining a to-be-screened sampleset according to relevance between sample features of any two training samples in a training sample set; screening a preset number of training samples from the to-be-screened sample set according tothe connection relationship between each training sample in the to-be-screened sample set and the adjacent training sample of each training sample, and generating a candidate sample set; wherein the adjacent training sample of each training sample is a training sample having a connection relationship with the training sample; determining a label fusion information entropy corresponding to each training sample according to each training sample in the candidate sample set and the adjacent training sample of each training sample; and screening out a target training sample for training the graph convolutional neural network from the candidate sample set according to the label fusion information entropy. According to the method, the performance obtained by the full data can be achieved by usinga small number of samples, and the calculation time consumption is reduced.
Description
technical field [0001] The embodiments of the present disclosure relate to the technical field of the Internet, and in particular to a training sample screening method, device, electronic equipment, and computer-readable storage medium. Background technique [0002] As a fast and effective means of man-machineverification, slider verification is a basic capability and a strong demand in anti-crawling, batch registration and other risk control fields. In addition to using artificial strategies for analysis, most algorithms use deep learning DNN (DeepNeural Network, deep neural network) and GNN (GraphNeural Network, graph neural network) to analyze and identify slider trajectory and motion characteristics. Recent research results show that GNN has higher accuracy and longer model degradation period than DNN. In order to achieve ideal performance, GNN needs to use a large number of labeled samples and predicted samples as nodes for map building. The time-consuming calculatio...
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
Application Date:The date an application was filed.
Publication Date:The date a patent or application was officially published.
First Publication Date:The earliest publication date of a patent with the same application number.
Issue Date:Publication date of the patent grant document.
PCT Entry Date:The Entry date of PCT National Phase.
Estimated Expiry Date:The statutory expiry date of a patent right according to the Patent Law, and it is the longest term of protection that the patent right can achieve without the termination of the patent right due to other reasons(Term extension factor has been taken into account ).
Invalid Date:Actual expiry date is based on effective date or publication date of legal transaction data of invalid patent.
Login to View More 
Login to View More