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

Integrated kernel density estimator window parameter optimization method, device and terminal device

A technology of kernel density estimation and window parameters, which is applied in the field of window parameter optimization of integrated kernel density estimators, can solve the problems of inaccurate probability density function estimation and high time complexity

Inactive Publication Date: 2018-12-21
SHENZHEN UNIV
View PDF0 Cites 4 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0004] The main purpose of the present invention is to propose an integrated kernel density estimator window parameter optimization method, device and terminal equipment to solve the inaccurate estimation of the probability density function in the prior art and the probability density of large-scale data sets When function estimation, the problem of high time complexity

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
  • Integrated kernel density estimator window parameter optimization method, device and terminal device
  • Integrated kernel density estimator window parameter optimization method, device and terminal device
  • Integrated kernel density estimator window parameter optimization method, device and terminal device

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0075] Such as figure 1 As shown, the embodiment of the present invention provides an integrated kernel density estimator window parameter optimization method, including the following steps:

[0076] S101. Obtain random sample division data blocks based on the original data set.

[0077] In the above step S101, the original data set includes one-dimensional data and multi-dimensional data, and the data blocks divided by random samples include one or more data in the original data set.

[0078] Such as figure 2 As shown, the detailed implementation process of obtaining random sample division data blocks in the above step S101 may include:

[0079] S1011. Divide the original samples of the original data set to obtain original sample division data blocks.

[0080] In the above step S1011, a plurality of original samples are selected in the original data set, and the data is divided where there are original samples, so as to obtain original sample division data blocks of the o...

Embodiment 2

[0118] The embodiment of the present invention takes a one-dimensional original data set as an example to illustrate the implementation process of the method for optimizing the window parameters of the integrated kernel density estimator provided in the first embodiment.

[0119] Assuming an existing dataset in is the number of training samples in the data set D.

[0120] Assume that a random sample of the existing data set D is divided into data blocks in To meet the conditions:

[0121] 1,

[0122] 2. For any And k i ≠k j , established;

[0123] E[p k (x)]=p(x), where p k (x) and p(x) respectively divide the data block D for random samples k And the probability density function of the original data set D, E(X) is the expectation of the random variable X.

[0124] Then obtain the random sample partition data block of the original data set D Specifically:

[0125] First, through the original data set D in the sample The formula for direct division ope...

Embodiment 3

[0151] The embodiment of the present invention aims at the window parameter optimization method of the integrated kernel density estimator provided in the first embodiment, and uses test data to exemplify the beneficial effect in its practical application.

[0152] The embodiments of the present invention respectively adopt the random number data sets of normal distribution and exponential distribution to verify the convergence and test error of the integrated kernel density estimator.

[0153] Five RSP data blocks are randomly selected from the random number data sets of normal distribution and exponential distribution, each data block contains 100 samples as the training set and verification set, and 200 random samples are randomly generated as the test set. The number of iterations of the particle swarm optimization algorithm in the integrated kernel density estimator is 200, and the initial window width parameter of the integrated kernel density estimator is randomly select...

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

A method for optimize window parameters of integrate kernel density estimator includes obtaining random sample to divide data block based on original data set, dividing data block by random sample, dividing data block by integrated kernel density estimator, dividing data block by integrated kernel density estimator, dividing data block by integrated kernel density estimator, dividing data block byintegrated kernel density estimator, dividing data block by integrated kernel density estimator, dividing data block by integrated kernel density estimator and dividing data block by integrated kernel density estimator. The integrated kernel density estimator is constructed on each random sample partition data block by Parzen window method, and the initial window width parameters of the integrated kernel density estimator are calibrated. K different random sample partition data blocks were selected to construct the training set; K different random sample partition data blocks were selected toconstruct verification sets; The initial window width parameters of the integrated kernel density estimator are optimized according to the training set and the verification set, and the optimal window width parameters are obtained. The integrated kernel density estimator is optimized according to the optimal window width parameter. The invention can improve the accuracy of the estimation of the probability density function, and can be applied to the estimation of the probability density function of a large-scale data set.

Description

technical field [0001] The invention relates to the field of data mining, in particular to a window parameter optimization method, device and terminal equipment of an integrated kernel density estimator. Background technique [0002] Probability density function estimation for data with unknown probability distribution is an important research content in the field of machine learning and data mining. The Parzen window method is a classic probability density function estimation method, also known as the kernel density estimation method. The key to estimating the probability density function of unknown probability distribution data using the kernel density estimation method is the selection of the window width parameter, which has a great influence on the kernel density estimator: too large a window width will lead to an over-smoothed Probability density function estimation; and too large window width will lead to undersmooth probability density function estimation. [0003]...

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 Applications(China)
IPC IPC(8): G06F17/30G06N3/00G06N7/00
CPCG06N3/006G06N7/01
Inventor 何玉林蒋捷黄哲学
Owner SHENZHEN UNIV
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