How to use a trained neural network in matlab??

I am not using neural net toolbox rather I am using my own code to build a neural net model. Now after saving my neural net result, I want to test it with new data set. How to do this???

Answer 1

[Ankit Bhardwaj]

Without knowing your custom neural network model, it is not possible for us to understand how it can be used to train new data set. However, you can basic workflow from the following example which shows how to classify new image data by fine tuning the existing, pretrained neural network.

https://www.mathworks.com/help/releases/R2017a/nnet/examples/transfer-learning-and-fine-tuning-of-convolutional-neural-networks.html

Since this is an example from Neural Network Toolbox, you need to have Neural Network Toolbox license to use some of the functions, but it can be used to learn some of the existing workflows.

[Hamza Ashraf]

hi i have tarined a network to detect specific sound now how can i use it. i want to provide audio coming from mike to network. can you tell me how to do that

Answer 2

[Greg Heath]

If you have saved the net, try

 ynew    = net( xnew );
 enew    = tnew - ynew;
 NMSEnew = mse( enew )/ mean( var( tnew',1 ))

Hope this helps.

Thank you for formally accepting my answer

Greg

[Hamza Ashraf]

hi i have tarined a network to detect specific sound now how can i use it. i want to provide audio coming from mike to network. can you tell me how to do that