The function call setLearnFunc is very similar to the setinitFunc call. setLearnFunc selects the learning function which will be used in the training process of the neural net. The format is:
setLearnFunc (function name, parameters....)
where function name is the name of the desired learning algorithm. This name is mandatory and has to match one of the following strings:
ART1 Dynamic_LVQ RadialBasisLearning ART2 Hebbian RBF-DDA ARTMAP JE_BP RCC BackPercolation JE_BP_Momentum RM_delta BackpropBatch JE_Quickprop Rprop BackpropMomentum JE_Rprop Sim_Ann_SS BackpropWeightDecay Kohonen Sim_Ann_WTA BPTT Monte-Carlo Sim_Ann_WWTA BBPTT PruningFeedForward Std_Backpropagation CC QPTT TimeDelayBackprop Counterpropagation Quickprop
After the name of the learning algorithm is provided, the user can specify some parameters. The interpreter is using default values if no parameters are selected. The values have to be of the type float. A detailed description can be found in the chapter ``Parameter of the learning function''. Function calls could look like this:
setLearn("Std_Backpropagation")
setLearn( "Std_Backpropagation", 0.1)
The first function call selects the learning algorithm and the second one additionally provides the first learning parameter. The batch interpreter displays:
Learning function is now: Std_backpropagation
Parameters are: 0.1