what is the difference between trainlm and trainbr?

hello matlab community, what is the difference between the Levenberg-Marquardt backpropagation (trainlm) and Bayesian Regularization (trainbr)??? Can anyone explain these two functions and the difference. Thank you for your help. :)

Answer 1

[Greg Heath]

One minimizes mean square error

The other minimizes a weighted sum of squared errors and squared weights

The latter is typically used to prevent overtraining an overfit net without using a validation subset.

Therefore it is most useful for small datasets to keep the training subset as large as feasible.

Hope this helps.

Thank you for formally accepting my answer

Greg