Hi, every one,
I bulid a shallow Neural network for predication of new design. Only have three features and one target.
I follow the workflow: https://www.mathworks.com/help/deeplearning/ug/multilayer-neural-networks-and-backpropagation-training.html
After train the dataset and evaluate the test data. A model (net) with acceptable rmse and high R value was consider as the best model.
I save the trained model(net) and use this model to predict.
However, I found that in some references or post, after determine the hyper-parameters,applying the chosen machine learning procedure on all of your data. which means use the entire dataset(train dataset + test dataset) to train the model again.https://machinelearningmastery.com/train-final-machine-learning-model/
the questions is : Do I also need finalize the model?
Best Answer