Using ANFIS for generating multiple outputs

I am trying to create an ANFIS model for processing 5 inputs and 4 outputs. It is actually a 4-class classification problem, where for each record, one output is 1(which is its class) and the rest are zero. The anfis() function does not support multiple outputs, but the genfis() function can accept multiple outputs (Documentation for genfis: Output data, specified as an M-column array, where M is the number of FIS outputs.). Once I have an initial FIS using genfis subtractive clustering, how do I proceed to train the model and evaluate it on test data?