4. (15 points) Given the artificial neural network, perform one training iteration using back-propagation for the input value of (0, 1) using XOR operation. The activation function to be used in the training is the sigmoid function. Initial values for all weights and thresholds will be taken as 0.5. Learning rate is 0.1. Momentum will not be employed. Show all your intermediate calculations with detailed equations in parts a, b and c. W13 x 1 W22 W35 es 5 W14 W45 x2 2 W24 te -1 Hidden Layer Input Layer Output Layer a. (5 points) Feed-forward phase: b. (5 points) Back-propagation phase: c. (5 points) Update phase