two questions about matlab online-course "Deep Learning with MATLAB"

I'm a student who want to learn deep learning with Matlab, so I take up the online course "Deep Learning with MATLAB". When I learn the chapter 5 ——"Image Classification Project", I got two questions which bothered me a lot.

1.How does the experiment manager know what plot to draw as a result? This case is a deep learning case, how does experiment manager know to draw the calculation process diagram and confusion matrix? I don't see the relevant Settings in the previous function Settings. Is it the default setting of experiment manager when performing deep learning calculations?

2.How to pass the best trained network to "bestExperimentNetworks"? I don't see any Settings for "bestExperimentNetworks "in the file. How are the training results passed to "bestExperimentNetworks"

Answer 1

[Taylor]

1. When you open the results of an experiment in Experiment Manager, look at the toolbar at the top. Ther eis a section called Review Results that has options for Training Plot, Training Data, and Validation Data (if you provided that). Select the one you want to view bly clicking on a trial in the table and the option you want from Review Results. Use the "plots" argument from the function trainingOptions to select the plots you want displayed in the Training Plot.
2. Select the result from the table you want to export. In the top-right of the toolbar, select "Export" to export the trained network. Expand the "Export" dropdown to access options to export training information and the results table as well.

[Zhizhuo]

thank u so much!!!