机器学习12

最新推荐文章于 2022-06-20 10:05:23 发布

lunei

最新推荐文章于 2022-06-20 10:05:23 发布

阅读量188

点赞数

分类专栏： Yanjiusheng

本文链接：https://blog.csdn.net/weixin_41865104/article/details/115742588

版权

Yanjiusheng 专栏收录该内容

32 篇文章 2 订阅

订阅专栏

#
# # softmax重点看看
# # 对于多分类问题实现概率和为 1 ,并且可以加大概率比较大的部分
#
#
#
# import torch
# '''# 交叉熵
#
# '''
import torch
import torch.nn.functional as F
from torch import optim, nn


from torchvision import datasets
from torchvision.transforms import transforms, Compose

x = torch.randn(1, 784)
w = torch.randn(10, 784)

logits = x @ w.t()
print(logits.shape)
print(logits)
pred = F.softmax(logits, dim=1)
# 归一化指数函数
print(pred.shape)
print(pred)
print(pred.sum())
pred_log = torch.log(pred)
print(pred_log)

print(F.cross_entropy(logits, torch.tensor([3])))
'''注意'''
# cross_entropy 运算 = pred = F.softmax(logits, dim=1) +torch.log(pred)否则使用以下
# 函数nll_loss
print(F.nll_loss(pred_log, torch.tensor([3])))
#

print('多分类案例')

w1, b1 = torch.randn(200, 784, requires_grad=True), torch.zeros(200, requires_grad=True)

print(b1)
w2, b2 = torch.randn(200, 200, requires_grad=True), torch.zeros(200, requires_grad=True)
w3, b3 = torch.randn(10, 200, requires_grad=True), torch.zeros(10, requires_grad=True)

torch.nn.init.kaiming_normal_(w1)
torch.nn.init.kaiming_normal_(w2)
torch.nn.init.kaiming_normal_(w3)
'''凯明初始化'''
def forward(x):
    x = x @ w1.t() + b1
    x = F.relu(x)
    '''第1层'''

    x = x @ w2.t() + b2
    x = F.relu(x)
    '''第2层'''

    x = x @ w3.t() + b3
    x = F.relu(x)
    '''第3层'''

    return x

batch_size =200
learning_rate = 10e-3
epochs = 10
train_loader = torch.utils.data.DataLoader(
    datasets.MNIST('./data', train= True, download=True, transform=Compose([transforms.ToTensor(),
                                                                            transforms.Normalize((0.1307,), (0.3081,))

                                                                                       ])),
    batch_size = batch_size, shuffle=True
)
test_loader = torch.utils.data.DataLoader(
    datasets.MNIST('./data', train= False, download=True, transform=Compose([transforms.ToTensor(),
                                                                            transforms.Normalize((0.1307,), (0.3081,))])),

    batch_size=batch_size, shuffle=True)


optimizer = optim.SGD([w1, b1, w2, b2, w3, b3], lr=learning_rate)
# 优化器,优化w1, b1, w2, b2, w3, b3

criteon = nn.CrossEntropyLoss()
# # 与cross_entropy 运算 = pred = F.softmax(logits, dim=1) +torch.log(pred)否则使用以下
for epoch in range(epochs):
    for batch_idx, (data, target) in enumerate(train_loader):
        data = data.view(-1, 28 * 28)
        logits = forward(data)
        loss = criteon(logits, target)

        optimizer.zero_grad()
        loss.backward()
        optimizer.step()
        if batch_idx % 100 ==0 :
            print('Train Epoch :{} [{}/{} ({:.0f}%)] \t Loss: {:.6f}'.format(epoch, batch_idx * len(data),
                                                                             len(train_loader),
                                                                          100. * batch_idx / len(train_loader),
                                                                          loss.item()))
    test_loss = 0
    correct = 0
    for data, target in test_loader:
        data = data.view(-1, 28*28)
        logits = forward(data)
        test_loss += criteon(logits, target).item()

        pred = logits.data.max(1)[1]
        correct += pred.eq(target.data).sum()

    test_loss /= len(test_loader.dataset)
    print('\nTest set: average loss : {:.4f}, Accuracy: {}/{} ({: .0f}%)\n'.format(test_loss, correct ,len(test_loader.dataset),

                                                                     100. * correct / len(test_loader.dataset)
                                                                     ))