simple_neural_network_part2.py

import numpy as np

inputLayerSize, hiddenLayerSize, outputLayerSize = 3, 3, 1
L = 0.25
X = np.array([[0, 0, 1], [0, 1, 1], [1, 0, 1], [0, 1, 0], [1, 0, 0], [0, 0, 0], [1, 1, 0]])
Y = np.array([[0], [0], [0], [0], [0], [0], [1]])
iterations = 50000

#sigmoid function
def f(x): return 1/(1 + np.exp(-x))
#derivative sigmoid
def f_(x): return x * (1 - x)

np.random.seed(1)
H1 = np.random.uniform(size=(inputLayerSize, hiddenLayerSize))
H2 = np.random.uniform(size=(hiddenLayerSize, outputLayerSize))

for i in range(iterations):
  R0 = np.dot(X, H1)
  R = f(R0)
  Oc = np.dot(R, H2)
  E = Y - Oc
  dOc = E * L
  dR = dOc.dot(H2.T) * f_(R)
  H2 += R.T.dot(dOc)
  H1 += X.T.dot(dR)

print "Input:\n[1, 1, 1]"
X = np.array([1, 1, 1])

print "\nOutput:"
print f(np.dot(X, H1)).dot(H2)