How does matlab normalize the mean square error?

This is probably very basic question. But I am very new to neural networks. So, I know in general mean square error (MSE) is calculated as MSE = where is the target value and is the neural network output. The "mse" function in matlab is called "mean square normalized error". So, my question is how and with what matlab is normalizing the error? Thank you.

Answer 1

[Alejandro Peñuelas]

This information is included with the documentation of mse function (here):

_______________________________________________________________________________________

This function has two optional parameters, which are associated with networks whose net.trainFcn is set to this function:

• 'regularization' can be set to any value between 0 and 1. The greater the regularization value, the more squared weights and biases are included in the performance calculation relative to errors. The default is 0, corresponding to no regularization.
• 'normalization' can be set to 'none' (the default); 'standard', which normalizes errors between -2 and 2, corresponding to normalizing outputs and targets between -1 and 1; and 'percent', which normalizes errors between -1 and 1. This feature is useful for networks with multi-element outputs. It ensures that the relative accuracy of output elements with differing target value ranges are treated as equally important, instead of prioritizing the relative accuracy of the output element with the largest target value range.

________________________________________________________________________________________

So, by default the function does not normalize the mse unless you indicate it in the parameters of the function.

Hope this can help you.

[Sai Praneet]

Thank you.

[Alejandro Peñuelas]

You're welcome.