undefined

1. Consider the following neural network: 24 23 y2 12 22 41 11 21 with the following weight matrices: W (10 5 3 4 2 5 8 6 1 2 8 3 W II [5 4 7 9] [2 3 9 4) So, for example, the weight between 22 and zz is 5 and the weight between 23 and y is 7. (a) Let x = [1, 2, 3] be an input to the network. Compute the outputs at y and y2 for this input. (10 pts) (b) Determine the partial error 8 for each of the nodes in the network w.r.t to z. (20 pts) (c) Update the weights in the network using gradient descent. Assume that the learning rate n = 1. (15 pts) 1. Consider the following neural network: 24 23 y2 12 22 41 11 21 with the following weight matrices: W (10 5 3 4 2 5 8 6 1 2 8 3 W II [5 4 7 9] [2 3 9 4) So, for example, the weight between 22 and zz is 5 and the weight between 23 and y is 7. (a) Let x = [1, 2, 3] be an input to the network. Compute the outputs at y and y2 for this input. (10 pts) (b) Determine the partial error 8 for each of the nodes in the network w.r.t to z. (20 pts) (c) Update the weights in the network using gradient descent. Assume that the learning rate n = 1. (15 pts)