how exactly to use the trainNetwork function?

I want to use the function net = trainNetwork(sequences,layers,options) like this to train a recurrent network of the form lstm for identifying nonlinear systems. I have for training two sets of input data and one set of output data.

numResponses=1;
featureDimension=1;
numHiddenUnits=70;
miniBatchSize=300;
maxEpochs=1000;
layer=[...
    sequenceInputLayer(featureDimension)
    lstmLayer(numHiddenUnits,'OutputMode','last')
    dropoutLayer(0.02)
    fullyConnectedLayer(numResponses)
    regressionLayer
    ];
options=trainingOptions('adam',...
    'MaxEpochs',maxEpochs,...
    'MiniBatchSize',miniBatchSize,...
    'GradientThreshold',20,...
    'Shuffle','once', ...
    'Plots','training-progress',...
    'ExecutionEnvironment','parallel',...
    'LearnRateSchedule','piecewise',...
    'LearnRateDropPeriod',200,...
    'L2Regularization',1e-3,...
    'LearnRateDropFactor',0.5,...
    'Verbose',0,...
    'Plots','training-progress');
C = num2cell(table2array(x1_train));
net = trainNetwork(C',layer,options);

Here is a part of my code.x1_train is a variable that contains the u1 regressor for the 2 tank system and I transformed it into a cell array because I understood that I had to do so that I could use this data in the train network function and I got the following error:

Error using trainNetwork

Not enough input arguments.

Error in sperproiectfinal1 (line 84)

net = trainNetwork(C',layer,options);

Caused by:

Error using nnet.internal.cnn.trainNetwork.DLTInputParser>iParseInputArguments

Not enough input arguments.

please help me with some steps that I should follow to use the function properly and be able to train my network with two sets of input data and one set of output data

Answer 1

[Cris LaPierre]

You need to include a response input.

net = trainNetwork(sequences,responses,layers,options)

From the documentation for sequence (Numeric or cell array row of the table) :

• If you specify sequences as a numeric or cell array, then you must also specify the responses argument.

[David Vatavu]

Hi Cris, thanks for your reply. But my question is: how do I find out variabla responses?

[Cris LaPierre]

It should be part of your training data set. You need to create a 'labeled' data set for training.

Responses.

When the input data is a numeric array or a cell array, specify the responses as one of the following.

• categorical vector of labels
• numeric array of numeric responses
• cell array of categorical or numeric sequences