Your first predictive model

Here's a simple training call to an YCML trainer, which returns a trained model, given existing input and output datasets:

YCFFN *theModel = [[YCRpropTrainer trainer] train:nil input:trainingInput output:trainingOutput];

YCML models and trainers may use YCMatrix instances in place of a dataframe. In such a case, YCML models accept matrices where each matrix column defines a single training example. Here is an example that uses matrices in place of Dataframes:

YCFFN *theModel = [[YCRpropTrainer trainer] train:nil inputMatrix:trainingInput outputMatrix:trainingOutput];

The resulting model may subsequently be used to make predictions, given a dataset or matrix:

YCDataframe *prediction = [theModel predict:testInput];