> For the complete documentation index, see [llms.txt](https://skywateryang.gitbook.io/timeseriesanalysis101/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://skywateryang.gitbook.io/timeseriesanalysis101/7.-ji-yu-shen-du-xue-xi-de-shi-jian-xu-lie-fen-xi-fang-fa/7.2-cnn-juan-ji-shen-jing-wang-luo.md).

# 7.2 CNN卷积神经网络

CNN计算机视觉领域占据着重要地位，而CNN同样可以用在时间序列上。区别在于应用在图像上的卷积核是二维的，而应用在时间序列上的卷积核是一维的，也就是一维卷积神经网络，1D CNN。

相比于基于RNN的LSTM等模型，1D CNN的优势是训练快，可以并行计算，并且在某些场景下可以获得不输给LSTM的模型效果。

下面就来学习如何用1D CNN训练时间序列数据。

```python
# 使用和上一节中LSTM准备好的相同数据样本，不再重复
# 构建一个简单的1D CNN模型
class CNNnetwork(nn.Module):
    def __init__(self):
        super().__init__()
        self.conv1d = nn.Conv1d(1,64,kernel_size=2)
        self.relu = nn.ReLU(inplace=True)
        self.fc1 = nn.Linear(64*11,50)
        self.fc2 = nn.Linear(50,1)

    def forward(self,x):
        # 该模型的网络结构为 一维卷积层 -> Relu层 -> Flatten -> 全连接层1 -> 全连接层2 
        x = self.conv1d(x)
        x = self.relu(x)
        x = x.view(-1)
        x = self.fc1(x)
        x = self.relu(x)
        x = self.fc2(x)
        return x
```

```python
torch.manual_seed(101)
model =CNNnetwork()

criterion = nn.MSELoss()
optimizer = torch.optim.Adam(model.parameters(), lr=0.001)

model
```

```python
# CNN用到的模型参数更少
def count_parameters(model):
    params = [p.numel() for p in model.parameters() if p.requires_grad]
    for item in params:
        print(f'{item:>6}')
    print(f'______\n{sum(params):>6}')

count_parameters(model)
```

```
   128
    64
 35200
    50
    50
     1
______
 35493
```

```python
epochs = 100
model.train()
start_time = time.time()

for epoch in range(epochs):

    for seq, y_train in train_data:

        # 每次更新参数前都梯度归零和初始化
        optimizer.zero_grad()

        # 注意这里要对样本进行reshape，转换成conv1d的input size（batch size, channel, series length）
        y_pred = model(seq.reshape(1,1,-1))
        loss = criterion(y_pred, y_train)
        loss.backward()
        optimizer.step()

    print(f'Epoch: {epoch+1:2} Loss: {loss.item():10.8f}')

print(f'\nDuration: {time.time() - start_time:.0f} seconds')
```

```python
future = 12

# 选取序列最后12个值开始预测
preds = train_norm[-window_size:].tolist()

# 设置成eval模式
model.eval()

# 循环的每一步表示向时间序列向后滑动一格
for i in range(future):
    seq = torch.FloatTensor(preds[-window_size:])
    with torch.no_grad():
        preds.append(model(seq.reshape(1,1,-1)).item())
```

```python
# 逆归一化还原真实值
true_predictions = scaler.inverse_transform(np.array(preds[window_size:]).reshape(-1, 1))
```

```python
# 对比真实值和预测值
plt.figure(figsize=(12,4))
plt.grid(True)
plt.plot(df['S4248SM144NCEN'])
x = np.arange('2018-02-01', '2019-02-01', dtype='datetime64[M]').astype('datetime64[D]')
plt.plot(x,true_predictions)
plt.show()
```

![](/files/-MiFhkO1wzk2txjDAjEX)

```python
# 放大看
fig = plt.figure(figsize=(12,4))
plt.grid(True)
fig.autofmt_xdate()

plt.plot(df['S4248SM144NCEN']['2017-01-01':])
plt.plot(x,true_predictions)
plt.show()
```

![](/files/-MiFhlV6P1HORdfjigQ6)