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

Neural network time sequence classification method based on data enhancement

A time series, neural network technology, applied in the field of time series classification, can solve the problem of insufficient training data set

Pending Publication Date: 2021-06-25
BEIJING UNIV OF TECH
View PDF0 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0005] The technical problem to be solved by the present invention is to provide a neural network time series classification method based on data enhancement to overcome the problem of insufficient training data sets

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
  • Neural network time sequence classification method based on data enhancement
  • Neural network time sequence classification method based on data enhancement
  • Neural network time sequence classification method based on data enhancement

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0036] The embodiment of the present invention provides a neural network time series classification method based on data enhancement. The present invention will be explained and elaborated below in conjunction with the relevant drawings:

[0037] On the basis of Mixup data enhancement, the present invention uses univariate time series data CinCECGTorso. The CinCECGTorso data set contains 4 categories, 1420 samples in total, and the sequence length is 1639. The LSTM-FCN network is used as the classification model.

[0038] Embodiment flow process of the present invention is as follows:

[0039] Step 1: For the CinCECGTorso time series dataset D={(x 1 ,y 1 ),(x 2 ,y 2 ),...,(x 1420 ,y 1420 )} for preprocessing; the specific steps include:

[0040] Step 1.1: Use the z-score standardization method to standardize the data set D, and the standardization formula is:

[0041]

[0042] where x i Indicates the i (1≤i≤1420) sample, μ indicates the sample mean, σ indicates the s...

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 discloses a neural network time sequence classification method based on data enhancement, and the method comprises the steps: carrying out the preprocessing of a time sequence data set; secondly, selecting a proper parameter alpha to carry out Mixup data enhancement on training data, wherein the enhanced data is used for model training; then, constructing an LSTM-FCN time sequence classification network model; and finally, taking a cross entropy function as a loss function, and training the LSTM-FCN network by using a back propagation and gradient descent algorithm Adam. According to the time sequence classification method based on data enhancement, the time sequence classification performance of the neural network is effectively improved.

Description

technical field [0001] The invention belongs to the field of time series classification, in particular to a neural network time series classification method based on data enhancement. Background technique [0002] Time series data widely exists in production and life, such as the trend of stocks, weather temperature, patient's electrocardiogram, etc. Analyzing these time series data and digging out important information is of great significance to guide people's production and life. Time series classification is an important and challenging task in time series problems. Traditional time series classification methods often rely on artificially designed features, and the calculation process is cumbersome and time-complex. The method based on deep learning requires a large amount of training data to ensure the generalization ability of the model to avoid overfitting and underfitting. This method is difficult to play a role when the amount of data is small. [0003] Current tim...

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): G16H50/70G06N3/04G06K9/62G06N3/08
CPCG16H50/70G06N3/084G06N3/044G06N3/045G06F18/214
Inventor 王天张婷刘兆英李玉鑑
Owner BEIJING UNIV OF TECH