Can the classifcation learner app and regression apps be used with sequential data?

3 views (last 30 days)
Sorry if this is a basic question but i find it hard to locate any details on what format data needs to be in to work with these apps in the documentation.
I have created a table containing 770 cells where each row has a single cell array in it (770X1). Each cell contained in the table has sequential data (8 rows x 1100 columns).
I added a 2nd column in my table that has the categorical label for each cell. So now my table is 770x2.
When i try to use the regression learner app i get an error saying there is 'no variable suitable as a response' and when i try to use the calssifcation learner app i get an error stating 'No predictor selected'
Is this becuase i am trying to use sequential data contained in cells or am i doing something else incorrect?

Accepted Answer

Shashank Gupta
Shashank Gupta on 24 May 2021
Hi Christopher,
Classification and regression learner APP does not take cell array as input, thus its not trivial to do a classification or regression on sequential data. Even if you are able to upload the data to learner APP, the existing model in the APP is not suitable for forecasting or dealing with sequencial data. Instead, a better choice would be use LSTM based network. I would recommend you to use Deep learning Designer APP. Check out this link, It will give you a good headstart.
I hope this helps.
Cheers
  1 Comment
CHRISTOPHER MILLAR
CHRISTOPHER MILLAR on 24 May 2021
HI,
Thanks for the reply. I thought that this was the case but was just checking. I have already developed an LSTM network i was just going to use the classification learner app as a way of comparing the results using the sequential data.
I have not used the Deep LEarning Designer though so thanks for pointing that out to me. It looks very easy to use
Regards
Chris

Sign in to comment.

More Answers (0)

Community Treasure Hunt

Find the treasures in MATLAB Central and discover how the community can help you!

Start Hunting!