Save and sort ANN models in each epoch and sort out the one with smallest MSE

Hi everyone.

I am trying to develop ANN models to predict a continuous target variable.

I want to randomly split my dataset and have access to each.

1. Import the train dataset and train the network ;
2. Import the validation dataset to validate the model;
3. import the test dataset to test the network.
4. Then compare it with measured values.

And also, I want to record every model generated during training, and sort out the best.

I tried youtube and google but they are not helping. Any advice or lead can help.

Note: Apologies if I fail to properly structure my question.

Thanks

Answer 1

[Bharat Chandra Mukkavalli]

Hi,

Please refer to this link on how to randomly split data into train, test and validation sets: https://in.mathworks.com/matlabcentral/answers/377839-split-training-data-and-testing-data

For model training, to save the model after every epoch refer to this link: https://in.mathworks.com/matlabcentral/answers/1573523-is-it-possible-to-save-results-after-every-epoch-so-that-if-there-is-some-issue-with-power-i-can-co