def gradient_check():
'''
梯度检查
'''
labels, data_set = transpose(train_data_set())
net = Network([8, 3, 8])
net.gradient_check(data_set[0], labels[0])
return net
def gradient_check():
'''
梯度检查
'''
labels, data_set = transpose(train_data_set())
net = Network([8, 3, 8])
net.gradient_check(data_set[0], labels[0])
return net
def test():
labels, data_set = transpose(train_data_set())
net = Network([8, 8, 8])
rate = 0.1
mini_batch = 100
epoch = 10
for i in range(epoch):
net.train(labels, data_set, rate, mini_batch)
print 'after epoch %d loss: %f' % (
(i + 1), net.loss(labels[-1], net.predict(data_set[-1])))
rate /= 2
correct_ratio(net)
def test():
labels, data_set = transpose(train_data_set())
net = Network([8, 8, 8])
rate = 0.3
epoch = 10
for i in range(epoch):
net.train(labels, data_set, rate, 10, 100)
print np.around(net.predict([data_set[-1]]), decimals=3).flatten()
print 'after epoch %d loss: %f' % (
(i + 1),
net.loss(labels[-1], net.predict([data_set[-1]]).reshape(8, 1))
)
rate /= 2
correct_ratio(net)
def test():
labels, data_set = transpose(train_data_set())
net = Network([8, 3, 8])
rate = 0.5
mini_batch = 20
epoch = 10
for i in range(epoch):
net.train(labels, data_set, rate, mini_batch)
print 'after epoch %d loss: %f' % (
(i + 1),
net.loss(labels[-1], net.predict(data_set[-1]))
)
rate /= 2
correct_ratio(net)
normalizer = Normalizer() correct = 0.0 for i in range(256) if normalizer.denorm(network.predict(normalizer.norm(i))) == i: correct += 1.0 print 'correct_ratio:' %.2f%%' % (correct / 256 *100) def test(): labels, data_set = transpose(train_data_set()) net = Network([8, 3, 8]) rate = 0.5 mini_batch = 20 epoch = 10 for i in range(epoch): net.train(labels, data_set, rate, mini_batch) print 'after epoch %d loss: %f' % ( (i + 1), net.loss(labels[-1], net.predict(data_set[-1])) ) rate /= 2 correct_ratio(net) def gradient_check() labels, data_set = transpose(train_data_set()) net = Network([8, 3, 8]) net.gradient_cheak(data_set[0], labels[0]) return net
