1. Consider the following Multi Layer Perceptron: 0-000000 1000000 (a) (2 pts) Let (1, y) E D be the sample retrieved from the dataset D, where I = {21, 12, 13] is the input and y {0,1} is the corresponding label. Let L be the Binary Cross En- tropy Loss function. What is the equation represented by this multi-layer perceptron? (You may write the sigmoid function as o in the equation) Solution: (b) (4 pts) Let w = 0.2, W2 = 1.4, W3 = -.5, W4 = 0.2 and x = [0.1, -1.1, -61". In the forward pass, calculate a, b, c, d, e, f and the loss L if y = 1. Use the natural logarithm In for all computations. Show your work. Solution: (c) (2 pts) Write down the chain-rule expansion for su Solution: (d) (4 pts) Calculate me You must show your work! Solution: 1. Consider the following Multi Layer Perceptron: 0-000000 1000000 (a) (2 pts) Let (1, y) E D be the sample retrieved from the dataset D, where I = {21, 12, 13] is the input and y {0,1} is the corresponding label. Let L be the Binary Cross En- tropy Loss function. What is the equation represented by this multi-layer perceptron? (You may write the sigmoid function as o in the equation) Solution: (b) (4 pts) Let w = 0.2, W2 = 1.4, W3 = -.5, W4 = 0.2 and x = [0.1, -1.1, -61". In the forward pass, calculate a, b, c, d, e, f and the loss L if y = 1. Use the natural logarithm In for all computations. Show your work. Solution: (c) (2 pts) Write down the chain-rule expansion for su Solution: (d) (4 pts) Calculate me You must show your work! Solution