6.2.2 Function Approximation from One to Two Dimensions

You can approximate functions with several outputs in the same way as with one output. Thus, you can proceed as in the previous example with the only difference being that the output data y should contain one output in each column. The basic approach is illustrated in Section 5.2.2, Function Approximation from One to Two Dimensions. To obtain an RBF example, you only have to change the initialization of the network to an RBF network and to reevaluate the whole example.