How to use the classification learner app?

Hi Guys,

I am working on a classification problem and not quite sure how I can use the classification learner app.

First of all, when I open the app and try to select the data from the workspace, the only data that shows up is the number of output classifiers which is 2:20.

For training data, I have 30624 cells that each of this cell contains a 3*600 matrix/double and the output label for each cell is an integer from 2 to 20.

So, the training set is a 1*30624 cell and the training label is a 1*30624 categorical.

I have the validation and test data set in the same format as training data.

Could someone please tell me how I should classify my data using the classification learner app? Any information would be appreciated.

Answer 1

[Ayush Gupta]

Type classification Learner in command line and classification learner app will open. Make sure to load the data in the workspace before opening the app. Select the variable as input and select the predictor and response variable. After that we can compare what all models we can use, and which will be better to use. Try to see the documentation on how to select from the various possible models we can use, here.

[Susan]

Hi Ayush,

Thank you very much for your response. The issue that I've been encountering is the loaded data in the workspace don't show up in the app that I can select them as an input/response variables. I thought maybe I need to reshape data in such a way that App can accept them. Any idea?

[Kuno Bruswachtl]

@Susan Hi,

I have the same problem where you cannot use cells as an Input. Any solutions to this?

[Ruben]

I have the same problem, is there still no answer or alternative solution? @Ayush Gupta @Kuno Bruswachtl@Susan

[Ayush]

Hi,

The Classification Learner app in MATLAB requires a tabular data set as input, where each row represents an observation and each column represents a predictor variable. The last column of the data set should contain the response variable, which is the variable to be predicted. The response variable should be categorical, numeric, or logical.

Additionally, the data set should be in one of the following formats:

- MATLAB table

- Numeric matrix

- Dataset array

- Datastore

The input data set can be loaded into the app from a workspace variable or from a file in a supported format such as CSV, XLSX, or MAT.

You can read more about it here Select Data for Classification or Open Saved App Session - MATLAB & Simulink (mathworks.com)

[Ruben]

@Ayush Thanks for rapid response,

I understand the kind of input requirements the Classification Learner has.

If you take a look at my original question "https://de.mathworks.com/matlabcentral/answers/1960204-how-can-i-import-a-cell-array-filled-with-tables-of-data-to-the-classification-learner-app" you might understand, what kind of solution I am looking for.

It would be nice if you could have a look at it, thank you!