利用深度卷积网络与小波变换实现对心电数据的学习分类

最新推荐文章于 2024-07-31 16:09:57 发布

MudFire

最新推荐文章于 2024-07-31 16:09:57 发布

阅读量2.9k

点赞数 4

分类专栏：学习神经网络作业

本文链接：https://blog.csdn.net/zhangbingxiazai/article/details/100064546

版权

该项目利用小波变换和深度卷积神经网络（CNN）对心电数据进行学习和分类。通过研究信号处理、图像处理的基本概念，实践使用预训练的GoogLeNet模型，调整以适应三个心电图类别，并训练新的网络。最终，从公共数据库下载数据，应用训练好的CNN进行分类并报告结果。

摘要由CSDN通过智能技术生成

Purpose：

1． This is the last project of this semester, finally!

2． This project will be involved with:

1) ECG data sets

2) Wavelet transform and scalogram

3) Convolution Neural Network

4) Some basic concepts of image processing

3．Please go to the following link and study “Signal Classification with Wavelet Analysis and Convolutional Neural Networks(CNN)*”

https://www.mathworks.com/help/wavelet/examples/signal-classification-with-wavelet-analysis-and-convolutional-neural-networks.html

*We will go over the concept of Convolution Neural Networks (CNN) in our next lesson.

4．Practice the examples demonstrated in the studying materials and configure a CNN for ECG classification.

5．Randomly download three ECG data sets from public ECG database, use the trained CNN to classify the downloaded data and report the results.

6．Due date: June 30, 2018.

BACKGROUND REIVEW：

Wavelet Analysis

Using continuous wavelet analysis, spectral features evolve over time can be explore and common time-varying patterns can be identified in two signals and perform time-localized filtering. Using discrete wavelet analysis, we can analyze signals and images at different resolutions to detect changepoints, discontinuities, and other events not readily visible in raw data. Signal statistics can be compared on multiple scales and hidden patterns will be revealed by performing fractal analysis of data.

Scalogram

In signal processing, a scalogram is a visual method of displaying a wavelet transform. There are 3 axes: x representing time, y representing scale, and z representing coefficient value. The z axis is often shown by varying the colour or brightness.

Convolution Neural Network

In machine learning, a convolutional neural network is a class of deep artificial neural networks, most commonly applied to analyzing visual imagery. Convolutional networks are inspired by biological processes in that the connectivity pattern between neurons resembles the organization of the animal visual cortex. Individual cortical neurons respond to stimuli only in a restricted region of the visual field known as the receptive field. The receptive fields of different neurons partially overlap such that they cover the entire visual field. Applications of CNN are developed for image and video recognition, recommender systems and natural language processing.

Convolution Neural Network

In machine learning, a convolutio