深度学习--全连接层、高阶应用、GPU加速
MSE均方差
Cross Entropy Loss:交叉熵损失
Entropy 熵:
1948年,香农将统计物理中熵的概念,引申到信道通信的过程中,从而开创了信息论这门学科,把信息中排除了冗余后的平均信息量称为“信息熵”。香农定义的“熵”又被称为香农熵或信息熵,即
其中标记概率空间中所有可能的样本,表示该样本的出现几率,是和单位选取相关的任意常数。
针对此问题,熵越大,不确定程度就越大,对于其中信息量的讨论参考知乎。
在信息学里信息量大代表着数据离散范围小,不确定性小。香农作为一个信息学家,他关心的是信息的正确传递,所以信息熵代表着信息传递的不确定性的大小。所以在信息学上,使用香农公式算出来的这个值,在信息学上叫做信息熵值,在熵权法中叫做冗余度值或者叫偏离度值,它的本来含义是指一个确定无疑的信息源发送出来的信息,受到干扰以后,衡量偏离了原始精确信息的程度。离散度越大,计算得这个值越小,则收到的信息越不可靠,得到的信息越小。这个值越大,则收到的信息越可靠,得到的信息越多。
在统计学里,就完全不是这样。统计学家不认为存在仅有一个的确定无疑的原始信息。而是认为收到的统计数字都是确信无疑的,只是由于发送主体可能是很多主体,或者是同一主体不同时间,不同地点,或者是统计渠道不同等等原因,得到了一组具有离散性的数值。在这种情况下,离散性越大,熵值越小,代表着信息量越大,所以权重越大。
a=torch.full([4],1/4)
#tensor([0.2500, 0.2500, 0.2500, 0.2500])
#计算交叉熵
-(a*torch.log2(a)).sum()
#tensor(2.)
交叉熵在神经网络中作为损失函数,p表示真实标记的分布,q则为训练后的模型的预测标记分布,交叉熵损失函数可以衡量p与q的相似性。交叉熵作为损失函数还有一个好处是使用sigmoid函数在梯度下降时能避免均方误差损失函数学习速率降低的问题,因为学习速率可以被输出的误差所控制。
交叉熵计算:H(p,q)=
MNIST再实现
import torch
import torch.nn as nn
import torch.nn.functional as F
import torch.optim as optim
from torchvision import datasets, transforms
batch_size=200
learning_rate=0.01
epochs=10
#加载数据集DataLoader(数据位置,batch_size,shuffle是否打乱,num_workers=4:4线程处理)
#torchvision.datasets.MNIST(root,train,transform,download) root指下载到的位置,train指是否下载训练集,transform指对图片进行转换后返回,download指是否下载
#torchvision.transforms([transforms.ToTensor(),transforms.Normalize((mean),(std))])
#transforms.ToTensor()做了三件事:1.归一化/255 2.数据类型转为torch.FloatTensor 3.shape(H,W,C)->(C,H,W)
#transforms.Normalize((mean),(std)) :用均值和标准差对张量图像进行归一化
train_loader = torch.utils.data.DataLoader(
datasets.MNIST('../data', train=True, download=True,
transform=transforms.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, transform=transforms.Compose([
transforms.ToTensor(),
transforms.Normalize((0.1307,), (0.3081,))
])),
batch_size=batch_size, shuffle=True)
w1, b1 = torch.randn(200, 784, requires_grad=True),\
torch.zeros(200, requires_grad=True)
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)
def forward(x):
x = [email protected]() + b1
x = F.relu(x)
x = [email protected]() + b2
x = F.relu(x)
x = [email protected]() + b3
x = F.relu(x)
return x
optimizer = optim.SGD([w1, b1, w2, b2, w3, b3], lr=learning_rate)
criteon = nn.CrossEntropyLoss()
for epoch in range(epochs):
for batch_idx, (data, target) in enumerate(train_loader):
data = data.view(-1, 28*28)
logits = forward(data)
# print(data.shape, target.shape,logits.shape)
loss = criteon(logits, target)
optimizer.zero_grad()
loss.backward()
# print(w1.grad.norm(), w2.grad.norm())
optimizer.step()
if batch_idx % 100 == 0:
print('Train Epoch: {} [{}/{} ({:.0f}%)]\tLoss: {:.6f}'.format(
epoch, batch_idx * len(data), len(train_loader.dataset),
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]
#print(pred)
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)))
全连接层
import torch
import torch.nn as nn
import torch.nn.functional as F
x=torch.randn(1,784)
x.shape
#torch.Size([1, 784])
# nn.Linear(输入、输出)
layer1 = nn.Linear(784,200)
layer2 = nn.Linear(200,200)
layer3 = nn.Linear(200,10)
x=layer1(x)
x=F.relu(x,inplace=True)
x.shape
#torch.Size([1, 200])
x=layer2(x)
x=F.relu(x,inplace=True)
x.shape
#torch.Size([1, 200])
x=layer3(x)
x=F.relu(x,inplace=True)
x.shape
#torch.Size([1, 10])
网络定义的高阶用法
import torch
import torch.nn as nn
import torch.nn.functional as F
import torch.optim as optim
class MLP(nn.Module):
def __init__(self):
super(MLP,self).__init__()
self.model = nn.Sequential(
nn.Linear(784,200),
nn.ReLU(inplace=True),
nn.Linear(200,200),
nn.ReLU(inplace=True),
nn.Linear(200,10),
nn.ReLU(inplace=True),
)
def forward(self,x):
x=self.model(x)
return x
net= MLP()
optimizer = optim.SGD(net.parameters(),lr=learning_rate)
criteon = nn.CrossEntropyLoss()
其他的激活函数 SELU、softplus、
GPU加速
import torch
import torch.nn as nn
import torch.nn.functional as F
import torch.optim as optim
from torchvision import datasets, transforms
batch_size=200
learning_rate=0.01
epochs=10
#加载数据集DataLoader(数据位置,batch_size,shuffle是否打乱,num_workers=4:4线程处理)
#torchvision.datasets.MNIST(root,train,transform,download) root指下载到的位置,train指是否下载训练集,transform指对图片进行转换后返回,download指是否下载
#torchvision.transforms([transforms.ToTensor(),transforms.Normalize((mean),(std))])
#transforms.ToTensor()做了三件事:1.归一化/255 2.数据类型转为torch.FloatTensor 3.shape(H,W,C)->(C,H,W)
#transforms.Normalize((mean),(std)) :用均值和标准差对张量图像进行归一化
train_loader = torch.utils.data.DataLoader(
datasets.MNIST('../data', train=True, download=True,
transform=transforms.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, transform=transforms.Compose([
transforms.ToTensor(),
transforms.Normalize((0.1307,), (0.3081,))
])),
batch_size=batch_size, shuffle=True)
class MLP(nn.Module):
def __init__(self):
super(MLP, self).__init__()
self.model = nn.Sequential(
nn.Linear(784, 200),
nn.LeakyReLU(inplace=True),
nn.Linear(200, 200),
nn.LeakyReLU(inplace=True),
nn.Linear(200, 10),
nn.LeakyReLU(inplace=True),
)
def forward(self,x):
x=self.model(x)
return x
##重点重点!!!
device=torch.device('cuda:0')
net = MLP().to(device)
optimizer = optim.SGD(net.parameters(),lr=learning_rate)
criteon = nn.CrossEntropyLoss().to(device)
for epoch in range(epochs):
for batch_idx, (data, target) in enumerate(train_loader):
data = data.view(-1, 28*28)
data,target = data.to(device),target.to(device)
logits = net(data)
# print(data.shape, target.shape,logits.shape)
loss = criteon(logits, target)
optimizer.zero_grad()
loss.backward()
# print(w1.grad.norm(), w2.grad.norm())
optimizer.step()
if batch_idx % 100 == 0:
print('Train Epoch: {} [{}/{} ({:.0f}%)]\tLoss: {:.6f}'.format(
epoch, batch_idx * len(data), len(train_loader.dataset),
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)
data, target = data.to(device), target.to(device)
logits = net(data)
test_loss += criteon(logits, target).item()
pred = logits.data.max(1)[1]
#print(pred)
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)))