Deep Learning Training and Validation Metrics

Hello,

I'm building a binary deep learning classification architecture using Matlab's Deep Learning Toolbox. Refering to the link:

https://au.mathworks.com/help/deeplearning/ref/trainingoptions.html

Where it shows an Image of "Training Progress", MATLAB uses 'Acuracy' as its metrics during training and validation (This metric is not used during training and validation, isn't it? Am I right?). The question would be: Is there a way to initialise the training option (or other parameters) such that it will use a different monitoring metrics such as 'precision', 'recall', 'f1 score' or a custom metric? If yes, is there a way to monitor multiple metrics?

Answer 1

[Divya Gaddipati]

Hi,

I have heard that this issue is known and the concerned parties may be investigating further.