Deep Learning and Neural Network Implementation

Suppose i have a trained network:

For example i have declared all other necessary training options and data to be used successfully.

The training was successfully done and i would like to use the "predict" comand on the test samples against the trained network "net"

net=trainNetwork(Xtrain,Xtarget,layers,opts);

% Prediction process

Y = predict(net,Xtest);

What will be data output on the variable ''Y'' ?

Is it probability scores?

Note: I have two classes for my training targets.

Answer 1

[Maksym Tymchenko]

The variable Y represents the predicted responses. The format and data type of the output Y depends on the type of problem, this can be:

• A numeric array
• A categorical array
• A cell array

For a detailed explanation see: https://www.mathworks.com/help/deeplearning/ref/seriesnetwork.predict.html#d124e144501

By default, for a classification problem, Y contains a matrix of prediction scores (probabilities). If you set the option ReturnCategorical to 1 (true), then the function returns categorical labels instead.