Dann backpropagate

Back to Dann

backpropagate( input , target , options );

Train the model's weights.

• input
  

Takes an array of input data.

• target
  

Takes an array of expected output.

• options (optional)
  

An object including specific properties.

Properties

Property	Type	Function
log	Boolean	If set to true, it will log a report in the console.
table	Boolean	If the 'log' option is set to true, setting this value to true will print the arrays of this function in tables.
saveLoss	Boolean	Whether or not to save the losses in the neural network object. After a lot of training, carrying loss data in the neural network object gets heavy, this is why it is set to false by default.
mode * for development	String	When gpu support will be implemented, specifing the string 'gpu' as opposed to 'cpu' will run the function on a kernel. This funtionality is not yet implemented

Example

//train 1 epoch
for (data of dataset) {
    nn.backpropagate(data.input,data.output);
}

or here is another example

//train 1 epoch
for (data of dataset) {
    nn.backpropagate([0,1],[1]);
    nn.backpropagate([1,0],[1]);
    nn.backpropagate([1,1],[0]);
    nn.backpropagate([0,0],[0]);
}