[0., 1., 1., 0., 1., 0., 0., 1., 0., 1., 1., 0., 1., 0., 0., 1.],
[8, 2])
data = nn.Dataset(x, y)
# 입력층 뉴련이 2개인 네트워크 생성
network = nn.Network(2)
# 뉴런이 4개인 히든층
network.append_affine(
4).append_batchnormalization().append_shift().append_sigmoid()
# 뉴런이 2개인 출력층
network.append_affine(2).append_softmax()
# 옵티마이저
optimizer = nn.optimizer.SGD(0.01)
# 학습을 위한 객체 생성
learner = nn.Learner(data, network, optimizer)
# 4개의 배치 사이즈로 내부 신경망을 초기화
learner.init(8)
# 처음 정확도 출력하기
print(learner.findAccuracy())
# 학습
learner.learn(1000, True)
# 결과
print(learner.findAccuracy())
import sys
sys.path.append(__file__.replace('\\NN\\example\\ex3_mnist1.py',''))
import NN as nn
import tensor
import pickle
network = nn.NetworkPlus(28*28)
network.append_affine(25).append_batchnormalization().append_relu()
network.append_affine(10).append_softmax()
optimizer = nn.optimizer.SGD()
with open(__file__.replace('ex3_mnist1.py', 'tensor_train_img.pkl'), 'rb') as f:
train_img = pickle.load(f)
with open(__file__.replace('ex3_mnist1.py', 'tensor_train_label.pkl'), 'rb') as f:
train_label = pickle.load(f)
train_img.shape = [train_img.shape[0], 28 * 28]
learner = nn.Learner(train_img, train_label, network, optimizer, 100)
learner.init()
learner.learn(10, True)
print(learner.findAccuracy())
