卷积神经网络09（多分类问题）太详细的，手把手

ont-hot 独热编码

使用e^z为了保证大于0

损失为0，对于分类是没有意义的

练习

import torch

criterion = torch.nn.CrossEntropyLoss()
Y = torch.LongTensor([2,0,1])
Y_pred1 = torch.tensor([[0.1,0.2,0.9],
                        [1.1,0.1,0.2],
                        [0.2,2.1,0.1]])

Y_pred2 = torch.tensor([[0.8,0.2,0.3],
                        [0.2,0.3,0.5],
                        [0.2,0.2,0.5]])

l1 = criterion(Y_pred1,Y)
l2 = criterion(Y_pred2,Y)
print("batch loss1 = ",l1.data,"\nbatch loss2 = ",l2.data)

batch loss1 = tensor(0.4966)
batch loss2 = tensor(1.2389)







最后一层不做激活

# import torch
#
# criterion = torch.nn.CrossEntropyLoss()
# Y = torch.LongTensor([2,0,1])
# Y_pred1 = torch.tensor([[0.1,0.2,0.9],
#                         [1.1,0.1,0.2],
#                         [0.2,2.1,0.1]])
#
# Y_pred2 = torch.tensor([[0.8,0.2,0.3],
#                         [0.2,0.3,0.5],
#                         [0.2,0.2,0.5]])
#
# l1 = criterion(Y_pred1,Y)
# l2 = criterion(Y_pred2,Y)
# print("batch loss1 = ",l1.data,"\nbatch loss2 = ",l2.data)

# ____________________________________________________________
# ____________________________________________________________
# ____________________________________________________________
import torch
from torchvision import transforms
from torchvision import datasets
from torch.utils.data import DataLoader
import torch.nn.functional as F
import torch.optim as optim
import matplotlib.pyplot as plt

batch_size = 64
transform = transforms.Compose([
    transforms.ToTensor(), #转成张量
    transforms.Normalize((0.1307,),(0.3081,)) # 神经网络使用(0,1)分布
])

train_dataset = datasets.MNIST('MNIST_data',
                               train=True,
                               download=True,
                               transform=transform)

train_loader = DataLoader(train_dataset,
                                shuffle = True,
                                batch_size = batch_size)

test_dataset = datasets.MNIST('MNIST_data',
                               train=False,
                               download=True,
                               transform=transform)

test_loader = DataLoader(test_dataset,
                                shuffle = False,
                                batch_size=batch_size)

# class DiabetesDataset(Dataset):
#     def __init__(self,filepath):
#         xy = np.loadtxt(filepath, delimiter=',', dtype=np.float32)  # delimiter分割 dtype=np.float32一般都使用32位的浮点数
#         self.len = xy.shape[0]  #可以取得输入的个数
#         self.x_data = torch.from_numpy(xy[:, :-1])
#         self.y_data = torch.from_numpy(xy[:, [-1]])  # [-1]拿出来的是一个矩阵而不是向量，要保证在计算的时候是矩阵
#
#     def __getitem__(self, index):   #魔方函数
#         return self.x_data[index],self.y_data[index]  #返回元祖类型，分开，便于后面使用和处理
#
#     def __len__(self):  #可以返回数据（数据的数量拿出来）
#         return self.len
#
# dataset = DiabetesDataset('diabetes.csv.gz')
#
# train_loader = DataLoader(
#     dataset = dataset,  # 数据集
#     batch_size= 32, # 小批量大小
#     shuffle=True, # 打乱
#     num_workers=2 # 是否需要多进程
# )


class Net(torch.nn.Module):
    def __init__(self):
        super(Net,self).__init__()
        self.l1 = torch.nn.Linear(784,512)
        self.l2 = torch.nn.Linear(512,256)
        self.l3 = torch.nn.Linear(256,128)
        self.l4 = torch.nn.Linear(128,64)
        self.l5 = torch.nn.Linear(64,10)

    def forward(self,x):
        x = x.view(-1,784)
        x = F.relu(self.l1(x))
        x = F.relu(self.l2(x))
        x = F.relu(self.l3(x))
        x = F.relu(self.l4(x))
        return self.l5(x)

model = Net()

criterion = torch.nn.CrossEntropyLoss()
optimizer = optim.SGD(model.parameters(),lr = 0.01,momentum=0.5)  # 带冲量的momentum

def train(epoch):
    running_loss = 0.0
    loss_1 = []
    epoch_l = []
    for batch_idx, data in enumerate(train_loader,0):
        inputs,target = data
        optimizer.zero_grad()

        outputs = model(inputs)
        loss = criterion(outputs,target)
        loss.backward()
        optimizer.step()
        # epoch_l.append(batch_idx)
        # loss_1.append(loss)

        running_loss += loss.item()
        if batch_idx % 300 == 299:
            print('[%d,%5d loss:%.3f'%(epoch + 1,batch_idx + 1,running_loss / 300))
            running_loss = 0.0
        # plt.plot(epoch_l, loss_1, c='r')
        # plt.show()


def test():
    correct = 0
    total = 0
    epoch_l = []
    loss_2 = []
    with torch.no_grad():  # 这里不会再做梯度计算
        for data in test_loader:
            images,labels = data
            outputs = model(images)
            _,predicted = torch.max(outputs.data,dim = 1)
            total += labels.size(0)
            correct += (predicted == labels).sum().item()
            epoch_l.append(epoch)
            correct = 100*correct/total
            loss_2.append(correct)
        return epoch_l,loss_2


    print('Accuracy on test set:%d %%'%(100*correct/total))

if __name__ == '__main__':
    a1= []
    b1 = []
    for epoch in range(10):
        train(epoch)
        a,b = test()
        a1.append(a)
        b1.append(b)

    plt.plot(a, b, c='r')
    plt.show()