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

A noise-tolerant online multi-classification kernel learning algorithm

A multi-classification and kernel learning technology, applied in computing, computer components, instruments, etc., can solve the problems that online learning algorithms cannot effectively control noise samples, and multi-classification methods cannot efficiently handle multi-classification of data streams, etc., to improve classification accuracy , good noise resistance, and reduced computational complexity

Active Publication Date: 2021-11-23
CHINA UNIV OF PETROLEUM (EAST CHINA)
View PDF5 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0003] The purpose of the present invention is to solve the problem that the existing multi-classification method based on batch processing technology cannot efficiently handle multi-classification of data streams, and the online learning algorithm cannot effectively control the influence of noise samples. Online Multi-Classification Kernel Learning Algorithm

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 noise-tolerant online multi-classification kernel learning algorithm
  • A noise-tolerant online multi-classification kernel learning algorithm
  • A noise-tolerant online multi-classification kernel learning algorithm

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0030] Embodiment 1: Take the online multi-classification experiments on the benchmark data sets AID7 data set, Outdoor Scene categories data set, UC Merced Landuse data set, and AID30 data set as examples for illustration. like figure 1 Shown is a schematic diagram of a noise-tolerant online multi-classification kernel learning algorithm provided according to an embodiment of the present invention. The online learning algorithm includes the following steps:

[0031] Step 1: Select an appropriate model kernel function and initialize the multi-classifier decision function. The specific steps are:

[0032] Specify the polynomial kernel function as the model kernel function, namely where the parameter γ is set to d is taken as the dimension of the sample x; c 0 Set to 0; polynomial degree p is set to 1. Initialize the multi-classification problem decision function f (0) =0.

[0033] Step 2: collect the data stream, and use the current decision function to predict the cat...

Embodiment 2

[0048] Embodiment 2: Different from Embodiment 1, in this embodiment, for the online learning algorithm based on kernel function, we use RBF kernel function where the parameter γ is set to d is the dimension of sample x.

[0049] The difference from Embodiment 1 is that in this embodiment, noise labels are added to the original benchmark dataset Adult data set, and an online classifier is trained on the dataset containing noise labels. Specifically, we will randomly select 5%, 10%, 15%, 20% (i.e. SNR 95:5, SNR 90:10, SNR 85:15, SNR 80:20) samples to change labels as noise data respectively.

[0050] image 3 The average test accuracy (ACA) comparison of the online classifiers Perceptron, Pegasos and the noise-resistant online multi-classification kernel learning algorithm based on the adaptive ramp loss function on the adult dataset containing noisy data. Experimental results show that our proposed noise-tolerant online multi-classification kernel based on the adaptive ramp...

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 relates to a noise-resistant online multi-classification kernel learning algorithm based on an adaptive ramp loss function. By introducing a kernel function to construct a nonlinear multi-classifier, aiming at the problem that the multi-classification method based on batch processing technology cannot efficiently deal with the data flow problem, and the existing online learning algorithm cannot effectively control the influence of noise samples, this noise-resistant online multi-classification method is designed. Kernel learning algorithm. This method can reduce the number of support vectors involved in model calculation, effectively control the influence of noise, significantly improve the efficiency of model update, improve the classification accuracy of multi-classification problems with noisy data, and meet the needs of practical application problems. The noise-resistant online multi-classification kernel learning algorithm of the present invention overcomes the problem that the traditional classification method based on batch processing technology cannot efficiently process data streams, and also overcomes the problem that existing online learning algorithms such as Perceptron and Pegasos cannot effectively suppress the influence of noise , which can be efficiently applied to practical application problems such as scene classification of pictures.

Description

technical field [0001] The invention belongs to the field of data mining and machine learning, and relates to data mining and data processing methods, in particular to a noise-resistant online multi-classification kernel learning algorithm based on an adaptive ramp loss function. Background technique [0002] The multi-classification problem is a classic problem in the field of data mining and machine learning. The traditional batch multi-classification method first collects data, builds a learning model based on a batch of data, and selects an optimization algorithm to obtain a multi-classifier. With the rapid development of technologies such as e-commerce, social media, mobile Internet, and the Internet of Things, more and more practical problems deal with data that has the characteristics of data streams. Traditional batch multi-classification methods have many shortcomings such as high computational complexity and low model update efficiency when dealing with large-scal...

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
CPCG06F18/2431G06F18/214
Inventor 渐令林丹周家欣宋允全高富豪雷鹤杰梁锡军
Owner CHINA UNIV OF PETROLEUM (EAST CHINA)
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