def compute_acc(X, Y, W):
y_hat = sr.compute_a(sr.compute_z(X, W, 0))
y_hat_indicies = np.argmax(y_hat, axis=1)
y_indicies = np.argmax(Y, axis=1)
error = np.array([y_indicies == y_hat_indicies])
acc = error.sum() / X.shape[0]
return acc
def compute_a2(z2):
'''
Compute the softmax activations a2 from the linear logits z2 in the second layer.
Input:
z2: linear logits in the second layer, a float numpy vector of shape c by 1.
Here c is the number of classes.
Output:
a2: the non-linear activations in the 2nd layer, a float numpy vector of shape c by 1.
Hint: you could solve this problem using 1 line of code.
'''
a2 = sr.compute_a(z2)
return a2
