Neural network App designer

Question

Cristina Barriopedro 2021-8-19

0
链接

此问题的直接链接

https://ww2.mathworks.cn/matlabcentral/answers/1437179-neural-network-app-designer

回答： Philip Brown 2023-5-4

I created an app that trains a neural network and then predicts different user inputs. I defined the net as app.net the whole training process, and I want to predict results with users input. I have problems with the predicting function, as the classify function gives me the following error, [YPred, scores] = classify(app.net1, User_NNPitstop'):

Error using classify (line 123)

Requires at least three arguments.

The predict function also gives me an error, and the test() gives NaN as a result.

The neural network whos net is class: network.

Thank you in advance!

0 个评论
显示 -2更早的评论隐藏 -2更早的评论

请先登录，再进行评论。

请先登录，再回答此问题。

Answer 1

Philip Brown 2023-5-4

0
链接

此回答的直接链接

https://ww2.mathworks.cn/matlabcentral/answers/1437179-neural-network-app-designer#answer_1229364

My guess is, instead of calling the classify() method of a deep neural network object, you are calling this classify function. That's likely happening because app.net1 is not the network object you expect it is.

Try using the debugger in AppDesigner to confirm the loaded network object is a deep neural network. It's possible that when you are importing the network (e.g. loading from file), you're not getting the network object you think you are.