how to classify with svm

HI
I have a number of photos that I have featured, that is, 11 properties for each photo. The number of photos is 770. I saved the data to an Excel file, which is 11 * 770.
How should I read this file?
And how do I train an SVM model using these features with svm (rbf, mlp, lvq)?
For svm, I need to take input from the Excel file.
Please help me soon.

1 Comment

To read the file you can use readtable(). Then follow the examples with the SVM. For example, you might try the "Classification Learner" app on the Apps tab of the tool ribbon. Attach your Excel workbook if you need more help and maybe someone will try something.

Sign in to comment.

Answers (1)

Bernhard Suhm
Bernhard Suhm on 1 Feb 2018
The table needs to have the various observations as rows. Alternatively to the ClassificationLearnerApp, you can train an SVM using the fitcsvm() function.
You can switch to a different kernel using the 'KernelFunction' input argument, for RBF it would be 'rbf'.
See https://www.mathworks.com/help/stats/fitcsvm.html for more details.

Categories

Asked:

on 28 Jan 2018

Answered:

on 1 Feb 2018

Community Treasure Hunt

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

Start Hunting!