Step 2: Training the Neural Network

>> mynet = nsTrain (mynet, x, y);

where, x is the input data and y is the desired data.

What do Input Data and Desired Data mean?

Input Data: Input data is the data that is fed to the neural network as input.
Desired Data: Desired data is the output you want the neural network to produce when given the corresponding input.

Let’s say you want the neural network to learn the sine function for values ranging from 0 to 10, then x would be

x =  0     0.1000     0.2000     0.3000     0.4000     0.5000  ...     9.7000     9.8000     9.9000    10.0000  

y would be the sine value of each element in x.

y =  0     0.0998     0.1987     0.2955     0.3894     0.4794     0.5646  ...    -0.2718    -0.3665    -0.4575    -0.5440  

Cross validation can be performed without any additional effort by passing the cross validation data to the train function as well.

>> mynet = nsTrain (mynet, x, y, cv_x, cv_y);

What is Cross Validation?

Cross validation tests the neural network on a cross validation data set at the same time that the network is being trained with the training data set. The weights of the network are saved when the error of the cross validation data set is at its minimum.

The learning curve and the output and desired plots can be seen with ease after training by setting their respective parameters to true.

>> mynet = nsnn;
>> mynet.learningCurve = true;
>> mynet.outputAndDesired = true;
>> mynet = nsTrain (mynet, inputData, desiredData);

Back | Tour Topics | Next