Is there any way to obtain the classified signal for each class ?

Question

OMAR MOUSA 2022-9-1

0
链接

此问题的直接链接

https://ww2.mathworks.cn/matlabcentral/answers/1792020-is-there-any-way-to-obtain-the-classified-signal-for-each-class

回答： Krishna 2023-12-27

Hi all,

I have four classes where they are electrical loads having current signal with respect to time ( sequential) which is one of the features

I am just wondering if there is any way to obtain the classified signal after classification because using a function (predict) just give me the classified labels of classes.

2 个评论
显示无隐藏无

Walter Roberson 2022-9-1

What difference do you see between giving the "classified signal after classification", compared to giving the "classified labels of classes" ?

OMAR MOUSA 2022-9-1

The difference is "classified signal after classification" the classified electrical signal for each class from the total power and "classified labels of classes" the classification result based on the labels

I mean, is there any way to have the classified signal with respect to time?

请先登录，再进行评论。

请先登录，再回答此问题。

Answer 1

Krishna 2023-12-27

0
链接

此回答的直接链接

https://ww2.mathworks.cn/matlabcentral/answers/1792020-is-there-any-way-to-obtain-the-classified-signal-for-each-class#answer_1378157

Hey Omar,

As far as I understand your question you need to classify a set of four signals and need a classified signal on the output’s labels. For “classified signals after classification” as you mentioned, can be created using the labels only, assuming you are able to classify the signal correctly. Given the question, the output signal for time k is nothing but the value of classified signal at time k. So, for example at the 10th timestep if you’re output label is 3 then the value of output signal is equal to value of signal 3 at timestep 10.

To correctly map 4 signals to classification label you can use several time series techniques some of them including RNN, LSTM in general. But for simple problems I would suggest starting with feed forward networks as base model as they can provide good results.

To know more about how to train classification model using feedforwardnet please go through this documentation link,

https://www.mathworks.com/help/deeplearning/ref/feedforwardnet.html

Note, these are shallow networks implementation. To create models using dlnetworks (latest network in MATLAB to implement deep learning models) please go through this documentation link which is an example for simple classifcation using dlnetworks,

https://www.mathworks.com/help/deeplearning/ug/create-simple-deep-learning-network-for-classification.html

Hope this helps.