Question: Apply an MLP network to solve the two-spirals problems (Figure) using in the previous chapter to illustrate the application of genetic programming. This data set is composed of 190 labeled samples (95 samples in each class), generated using polar coordinates. Determine a suitable number of neurons in the MLP network to be trained using the standard backpropagation algorithm. Assume the algorithm has converged when MSE ≤ 10-3, and plot the decision surface generated separating both classes. Test the sensitivity of the trained network to various levels of noise added to the input data.