An on-line soft-spaced kernel learning algorithm based on step size control

A soft interval and kernel learning technology, applied in the field of online soft interval kernel learning algorithm, can solve the problem that the classification method cannot efficiently handle the data stream classification problem, and the online learning algorithm cannot suppress the influence of noise, etc., so as to improve the classification accuracy and reduce the computational complexity. degree, the effect of reducing the running time

Active Publication Date: 2019-01-25
CHINA UNIV OF PETROLEUM (EAST CHINA)
View PDF6 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 propose an online soft interval kernel learning algorithm based on step size control, aiming at the fact that the existing classification method based on batch processing technology cannot efficiently handle the data stream classification prob

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
  • An on-line soft-spaced kernel learning algorithm based on step size control
  • An on-line soft-spaced kernel learning algorithm based on step size control
  • An on-line soft-spaced kernel learning algorithm based on step size control

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0033] Embodiment 1: Take the online classification experiment on the original benchmark data sets ijcnn, codrna, and eegeye as an example for illustration. like figure 1 Shown is a schematic diagram of an online soft interval kernel learning algorithm based on step size control provided according to an embodiment of the present invention. The online learning algorithm includes the following steps:

[0034] Step 1: Initialize model parameters, decision function and model kernel function. The specific steps are:

[0035] Initialize the model threshold parameter C=0.05, initialize the decision function f of the binary classification problem 0 =0, specify the Gaussian kernel function as the model kernel function, namely k(x i ,x j )=exp(-‖x i -x j ‖ 2 / d), where d is taken as the dimensionality of the sample input x.

[0036] Step 2: collect the data stream, and use the classification decision function to predict the category label of the data stream sample. The specific...

Embodiment 2

[0046] Embodiment 2: On the basis of the original benchmark data sets ijcnn, codrna, and eegeye, noise labels are added, and an online classifier is trained on the data sets containing noise labels. The difference from Embodiment 1 is that in this embodiment, in Step 1, 30% of the data set is randomly selected as a test set, and the rest of the data is added to noise labels to construct a training set. Specifically, we respectively modulo 20, modulo 10, and modulo 5 the sample indices, and multiply the sample point labels with a remainder of 0 by -1 to obtain the noise label data.

[0047] Figure 3-5 For training online classifiers KernelPerceptron, Pegasos and OSKL on datasets ijcnn, codrna and eegeye with noisy labels, and the average classification performance (average test accuracy, ACA) on the original 30% dataset without noise test dataset. The experimental results show that as the noise of the training samples indexed by mod20, mod10 and mod5 increases, the classifica...

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 an on-line soft-spaced kernel learning algorithm (OSKL) based on step size control. By introducing kernel function to construct nonlinear classifier and soft interval parameter to control the effect of noise data, a robust on-line kernel learning algorithm is designed based on the basic framework of on-line gradient descent algorithm. The algorithm can reduce the storage space of the model and effectively control the influence of noise. The computational complexity of model updating is only O (1). It is a natural tool to deal with and analyze the data flow problems because of its advantages of real-time and easy implementation. The online learning algorithm of the invention overcomes the problem that the traditional classification method based on batch processing technology can not efficiently process the data stream, and also overcomes the problem that the existing online learning algorithm such as Kernel Perceptron and Pegasos and the like can not effectivelysuppress the influence of noise.

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 an online soft interval kernel learning algorithm (OSKL) based on step size control. Background technique [0002] Classification problem is a classic research problem in the field of data mining and machine learning. The traditional classification method based on batch processing technology first collects data, builds a learning model based on the collected data, and selects an optimization algorithm to solve the model to obtain a classifier. With the rapid development of e-commerce, social media, mobile Internet, Internet of Things and other technologies, more and more application scenarios need to process large-scale data streams in real time. Traditional classification methods based on batch processing technology have many shortcomings such as high computational complexity and low model update efficiency ...

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): G06K9/62
CPCG06F18/214G06F18/24
Inventor 宋允全雷鹤杰吕聪梁锡军渐令
Owner CHINA UNIV OF PETROLEUM (EAST CHINA)
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