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

Feature selection algorithm based on ReliefF-DDC

A feature selection and algorithm technology, applied in computing, computer components, climate sustainability, etc., can solve the problem of insufficient load feature selection, and achieve the effect of increasing the load and reducing the feature dimension.

Active Publication Date: 2020-11-06
NANJING INST OF TECH
View PDF2 Cites 2 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

In contrast, the research on load feature selection is slightly insufficient

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
  • Feature selection algorithm based on ReliefF-DDC
  • Feature selection algorithm based on ReliefF-DDC
  • Feature selection algorithm based on ReliefF-DDC

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0037] The present invention will be specifically described below in conjunction with the accompanying drawings.

[0038] as attached figure 1 As shown in the flowchart, a feature selection algorithm based on ReliefF-DDC is characterized in that, comprising the following steps:

[0039] Step 1: Obtain training set samples and determine the parameter values ​​involved in the algorithm; specifically, the following steps are included:

[0040] S11. Let the training set to be processed be D, and the sample sample X l ={x l1 , x l2 ,...x ld},x ld is the d-dimensional feature of the l-th sample in the training set D.

[0041] S12. Determine the number of iterations m, m≥1; feature weight threshold τ, 0≤τ≤1; the number of nearest neighbor samples k, k is an integer greater than or equal to 1; evaluation criterion threshold δ, 0≤δ≤1.

[0042] Step 2: Reset all feature weights in the sample to 0, and set F and S as empty sets.

[0043] Step 3: Select sample X from training set ...

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 particularly relates to a feature selection algorithm based on ReliefF-DDC, and the algorithm comprises the following steps: S1, obtaining a training set sample, and determining each parameter value of the algorithm; S2, setting all feature weights to be 0, and setting the weights as a null set; s3, selecting a sample from the training set, updating the weights of the features of alldimensions contained in the sample, calculating the correlation between the features and each category by using ReliefF to determine important features, and excluding irrelevant features; s4, outputting the corresponding feature vectors when the feature vectors are greater than the threshold value, and adding the feature vectors to the set in a descending order; s5, using a DDC algorithm to analyze and remove redundant features according to the correlation between the features and the decision variables; and S6, obtaining an optimal feature subset, and applying the selected features to non-intrusive load identification. According to the method, the feature dimension is effectively reduced, the load identification rate is improved and the algorithm operation time is shortened.

Description

technical field [0001] The invention relates to a non-intrusive load feature selection algorithm, in particular to a feature selection algorithm based on ReliefF-DDC. Background technique [0002] Non-intrusive Load Monitoring (NILM) provides data support for the interaction between the smart grid and users. This method installs sensors at the entrance of the household line to collect electrical loads such as voltage and current of the total load. Analyze the volume data and refine the system data to identify the category and operating status of household appliances. Compared with the intrusive load monitoring method (ILM), NILM has the advantages of low cost, high user acceptance, and convenient maintenance in the later stage, but this method has higher requirements for the load decomposition algorithm. Feature extraction and load recognition, as two key technologies in NILM, provide strong technical support for the development of NILM. [0003] At present, most of the re...

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): G06K9/62
CPCG06F18/24143G06F18/214Y04S20/242
Inventor 邵琪包永强贾成宇张旭旭陆志文
Owner NANJING INST OF TECH