Cross -validation implemented in the the nprtool

1 view (last 30 days)
Yogini Prabhu
Yogini Prabhu on 24 May 2021
Answered: SM on 12 Jan 2022
Please let know how is Cross -validation implemented in the the code of classification of patterns at nprtool (Deep Learning Toolbox)?
  4 Comments
John D'Errico
John D'Errico on 30 May 2021
Why ask the exact same question twice? PLEASE DON"T DO IT A THIRD TIME.
Yogini Prabhu
Yogini Prabhu on 3 Jun 2021
Hi John,
I am sorry for repeating my question multiple times,
I was in urgent need of the query resolution,
and at time of submitting the Q, I might of encountered network error.
so I might have repeared the action such that at least one of Q is uploaded.
However, capitals is shouting (now as per rules of Q&A too).

Sign in to comment.

Answers (3)

Rishik Ramena
Rishik Ramena on 26 May 2021
Edited: Rishik Ramena on 27 May 2021
If you want to check the implementations of uncompiled matlab functions, use the type command from the MATLAB command window. For the nprtool,
type nprtool
But this wouldn't work for compiled in-built functions.
  3 Comments
Walter Roberson
Walter Roberson on 29 May 2021
Shallow neural networks do not exist in the Statistics and Machine Learning toolbox; they are only in the Deep Learning Toolbox (which used to be named the Neural Network toolbox)
Yogini Prabhu
Yogini Prabhu on 30 May 2021
okay! question is corrected as above with accuracy.
what about the answer?

Sign in to comment.


Walter Roberson
Walter Roberson on 4 Jun 2021
nprtool uses traincsg(), which does not do cross-validation.
Therefore the answer to how Cross-Validation (CV) is done with nprtool is that CV is not done for nprtool.

SM
SM on 12 Jan 2022
Hi,
I think nprtool default cross-validation techniuqe is "hold out". It seems that for k-fold cross validation you should export the script and modify it.
Have a look at this link, you may find it useful.

Tags

Community Treasure Hunt

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

Start Hunting!