RBF (Radial Basis Function)

MLP (Multi Layer Perceptron)

Part 3: Fmction Approximation by Neural Networks (i) For the x-y fimction in Figure 1, approximate the fimction using a 3-layer MLP with sigmoid activation fumnctions in the hidden layer and linear fimctions in the ourput layer. Show the neural network structure graphically, and indicate the approximate values of neural network internal biases and connection weights Figure 1 (ii). For thex-y fimction in Figure 2, approximate the fimction using a RBF network with Gaussian activation fumctions in the hidden layer and linear finctions in the output layer. Show the RBF neural network structure graphically, and indicate the approximate values of neural network intermal parameters including parameters for the Gaussian fmctions and the neural network coanection weights 15 10 15 20 25 -5 Figure 2 Part 3: Fmction Approximation by Neural Networks (i) For the x-y fimction in Figure 1, approximate the fimction using a 3-layer MLP with sigmoid activation fumnctions in the hidden layer and linear fimctions in the ourput layer. Show the neural network structure graphically, and indicate the approximate values of neural network internal biases and connection weights Figure 1 (ii). For thex-y fimction in Figure 2, approximate the fimction using a RBF network with Gaussian activation fumctions in the hidden layer and linear finctions in the output layer. Show the RBF neural network structure graphically, and indicate the approximate values of neural network intermal parameters including parameters for the Gaussian fmctions and the neural network coanection weights 15 10 15 20 25 -5 Figure 2