def test_feed_forward(self):
xor_network = [ # hidden layer
[[20, 20, -30], [20, 20, -10]],
# output layer
[[-60, 60, -30]],
]
output = []
for x in [0, 1]:
for y in [0, 1]:
print x, y, neural.feed_forward(xor_network, [x, y])[-1]
output.append(neural.feed_forward(xor_network, [x, y])[-1])
print output
def predict(input):
return is_neural.feed_forward(network, input)[-1]
"""return a matplotlib 'patch' object with the specified
location, crosshatch pattern and color"""
return mpl.patches.Rectangle((x-0.5, y-0.5), 1, 1,
hatch=hatch, fill=False, color=color)
for i in range(5):
for j in range(5):
if weights[5*i + j] < 0:
# add black and white hatches, so visible whether dark or light
ax.add_patch(patch(j, i, '/', "white"))
ax.add_patch(patch(j, i, '\\', "black"))
# plt.show()
left_column_only = [1, 0, 0, 0, 0] * 5
print is_neural.feed_forward(network, left_column_only)[0][0] # 10
center_middle_row = [0, 0, 0, 0, 0] * 2 + [0, 1, 1, 1, 0] + [0, 0, 0, 0, 0] * 2
print feed_forward(network, center_middle_row)[0][0] # 0.95
right_column_only = [0, 0, 0, 0, 0] * 5
print feed_forward(network, right_column_only)[0][0] # 0.0
my_three = [
0,1,1,1,0, # .@@@
0,0,0,1,1, # ...@@
0,0,1,1,0, # ..@@
0,0,0,1,1, # ...@@
0,1,1,1,0] # .@@@.
sigmoid(.121 * -11.61 + 1 * -2.17 + 1 * 9.31 - 1.38 * 1 - 0 * 11.47 - 1.92)
