Supercharge Your Innovation With Domain-Expert AI Agents!

Concept drift detection method based on divergence and EWMA

A concept drift and detection method technology, which is applied in the field of concept drift detection based on divergence and EWMA, can solve the problems of reducing the accuracy of classifiers, achieve the effect of solving the problem of concept drift and the decline of classification accuracy, and improving the classification results

Pending Publication Date: 2022-07-29
NANJING UNIV OF AERONAUTICS & ASTRONAUTICS
View PDF0 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

This change is reflected in the incoming instances and reduces the accuracy of the classifier learned from past training instances

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
  • Concept drift detection method based on divergence and EWMA
  • Concept drift detection method based on divergence and EWMA
  • Concept drift detection method based on divergence and EWMA

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0021] The present invention will be further described below with reference to the accompanying drawings and specific embodiments.

[0022] combine figure 1 , the present invention proposes a concept drift detection method based on divergence and EWMA, and the specific implementation steps are as follows:

[0023] Step 1: For example, set the window data size to 500, divide the data stream into the window according to the initial set size, initially use the first window as a fixed window, continue to divide the subsequent data into sliding windows, and divide the sliding window and the fixed window. The data in the window is mapped to the data distribution of the corresponding window in the form of data frequency, and its calculation formula is where N(x|w) represents the number of feature vectors x in the window w, and n represents the window size. The sliding window model is attached figure 2 As shown, when there is no concept drift, only the second window is slid. When...

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 divergence and EWMA-based concept drift detection method, which aims at detecting whether a data stream generates concept drift or not by measuring data distribution differences between divided sliding windows, and comprises the following steps of: 1, dividing the sliding windows from the data stream, and constructing a data distribution function of the windows based on window data; and 2, the difference of data distribution between the sliding windows is measured by using the Jensen-Shannon divergence. And 3, judging whether concept drift is generated or not through an EWMA (Exponentially Weighted Moving Average) hypothesis testing mode, and retraining a new classifier to adapt to subsequent data after the concept drift is detected.

Description

technical field [0001] The invention belongs to the field of data stream processing, in particular to an unsupervised online concept drift detection method based on divergence and EWMA (Exponential Weighted Moving Average). Background technique [0002] The data processed in the past tended to be static data that could be stored in memory and processed on entire datasets. However, with the rapid development of information technology, data has appeared in the form of a stream of continuous arrival. Compared with traditional data, the data stream has a large amount of data and arrives in real time, and once the data is processed, it cannot be taken out for processing unless the data is deliberately saved. Data in real-world environments may have dynamic behavior and concepts change, which is the so-called concept drift problem. [0003] The definition of concept drift is that in a given time period [0, t], the data flow in this time period is denoted as S 0，t ={d 0 ,...,d ...

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): G06F16/2455G06K9/62G06N3/08
CPCG06F16/24568G06N3/088G06F18/24G06F18/214
Inventor 赵蕴龙范其林
Owner NANJING UNIV OF AERONAUTICS & ASTRONAUTICS
Features
  • R&D
  • Intellectual Property
  • Life Sciences
  • Materials
  • Tech Scout
Why Patsnap Eureka
  • Unparalleled Data Quality
  • Higher Quality Content
  • 60% Fewer Hallucinations
Social media
Patsnap Eureka Blog
Learn More

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

© 2025 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap|About US| Contact US: help@patsnap.com