Hi Berna,
I understand that you want to know the type of performance in the GUI of nntraintool and perfom() function in MATLAB.
The performance value shown in the `nntraintool` window represents the training performance of your network. It is the performance achieved on the training data during the training process. In your case, the value of 0.0473 indicates the training performance achieved by your network.
On the other hand, the `performance = perform(net,t,y)` line in your code calculates the performance of the network on a specific set of data. The `t` and `y` variables represent the target outputs and the network's predicted outputs, respectively. By calling `perform(net,t,y)`, you can calculate the performance of the network on this specific dataset.
It is possible that the performance calculated using `perform(net,t,y)` may differ from the performance shown in the `nntraintool` window. It can be due to the following reasons:
The performance in the `nntraintool` window is typically calculated using the training data, while `perform(net,t,y)` allows you to evaluate the performance on any dataset, including training, validation, or test data.
Therefore, if you want to compare the performance values obtained from the `nntraintool` window and the `perform` function, make sure you are evaluating them on the same dataset. If you use the training data for both, the values should be similar. However, if you calculate the performance on a different dataset, such as validation or test data, the results may differ due to the network's generalization ability.
Note: The ‘nntraintool’ function was removed in the R2021b Matlab release. It is recommended to use the ‘train’ function.
For more information on ‘nntraintool’, you can refer to MATLAB documentation for R2021a here https://www.mathworks.com/help/releases/R2021a/deeplearning/ref/nntraintool.html?searchHighlight=nntraintool&s_tid=doc_srchtitle.
For more information on ‘perform()’, you can refer to the MATLAB documentation here https://www.mathworks.com/help/deeplearning/ref/network.perform.html?s_tid=doc_ta.
