Unlock instant, AI-driven research and patent intelligence for your innovation.

Method for reducing training time and supporting vector

A support vector and training time technology, applied in the field of information processing, can solve the problems of long training time and many support vectors, and achieve the effect of streamlining training samples, reducing support vectors, and reducing training samples

Inactive Publication Date: 2008-06-25
SHANGHAI JIAO TONG UNIV
View PDF0 Cites 2 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0004] The present invention aims at above-mentioned deficiencies in the prior art, proposes a kind of method of reducing training time and support vector, makes it solve the deficiency that existing support vector machine method is too long in training time, too many support vectors when solving large-scale problem

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

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0026] The data used in this embodiment is the Banana (banana) database provided by Benchmark (reference database), wherein the first 1-50 groups of training samples are used, each group of sample size is 400, a total of 20000 training samples, the front of the test sample For groups 1-5, the number of samples in each group is 4900, a total of 24500 test samples.

[0027] Step 1, extracting neighboring samples from the training samples to obtain a neighboring sample set, and obtain the boundary information of the spatial distribution.

[0028] Step 1. If there is only one two-class sample set in the training sample, the two-class sample set includes: positive class samples and negative class samples, calculate the distance from each sample in one class to each sample in the other class, and each The distance corresponds to two samples belonging to two categories; if the training sample exceeds one two-type sample set, multiple two-type sample sets are synthesized by two groups...

Embodiment 2

[0050] The database of this embodiment is the Waveform database provided by Benchmark, which is two types of samples, and the input dimension is 21. The database has a total of 100 training sample sets and 100 test sample sets each, 400 samples in each training sample set, and 4600 samples in each test sample set. In this embodiment, a total of 10,000 samples of the first 1-25 groups of the training sample set are used as training samples, and a total of 9,200 samples of the first 1-2 groups of the test sample set are used as test samples.

[0051] Table 2 The comparison after simplification with the optimal parameters under the original sample

[0052] Reduced Radius

[0053] The specific operation process of this embodiment is the same as that of Embodiment 1, and no further elaboration will be made here. However, different from Embodiment 1, the parameters for training the support vector machine in this embodiment are not deliberately selected, and are Gamma...

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 provides a method for reducing the training time and supporting vectors, belonging to the intelligent information processing technical field. The method comprises the following steps that; Step 1. a critical sample is extracted from training samples to get a critical sample set so as to get the boundary information of spatial distribution; Step 2. after the critical sample is extracted in the step 1, a non-critical sample is extracted from training samples to get a simplified sample set; Step 3. the critical sample set and the simplified sample set are united to get an ultimate training sample set. Because the invention saves the boundary properties of the sample distribution and the non-critical sample and use the ultimate training sample set to support the training of a vector machine so as to get an ultimate classifier, the invention simplifies the training samples to a great extent and has a mostly unchanged generalization ability.

Description

technical field [0001] The invention relates to a method in the technical field of information processing, in particular to a method for reducing training time and support vectors. Background technique [0002] People begin to understand the world by classification, which is the most basic way for people to understand the world. The traditional classification method kNN (K-Nearest Neighbor) is a basic and important method. With the expansion of the scope of neural network applications and the increase of application complexity, several new methods with better performance were proposed in the 1990s. The most representative and independent system method is the M 3 method (Max-Min-Model maximum and minimum module) and the representative method SVM (Support Vector Machine) proposed by Vapnik on structural risk minimization. The SVM method is the implementation method of the structural risk minimization theory. Its main idea is to solve the maximum distance between two types o...

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): G06F17/30G06F15/18
Inventor 陈玉坤
Owner SHANGHAI JIAO TONG UNIV