O: R+R, o(a) = { Problem 5. (10+5 marks] In this whole exercise we use neurons with respect to the Heaviside activation function given by 1 for r>0 0 otherwise. (i) Carry out the perceptron algorithm to approximate the logical AND operator where weights and bias are initialized as wj = 1, 2 = 2 and b = -1. (ii) Can the XOR logical operator be represented by a single neuron, too? If not, are there alternatives to single neurons? Hint: For (ii) it is not necessary to give any proof. O: R+R, o(a) = { Problem 5. (10+5 marks] In this whole exercise we use neurons with respect to the Heaviside activation function given by 1 for r>0 0 otherwise. (i) Carry out the perceptron algorithm to approximate the logical AND operator where weights and bias are initialized as wj = 1, 2 = 2 and b = -1. (ii) Can the XOR logical operator be represented by a single neuron, too? If not, are there alternatives to single neurons? Hint: For (ii) it is not necessary to give any proof