当前位置：首页 > wzjs >正文

衢州企业网站建设房地产估价师考试

wzjs 2025/8/14 5:37:42

衢州企业网站建设,房地产估价师考试,个人网站的开发与设计,做网站什么费用🍨 本文为🔗365天深度学习训练营中的学习记录博客🍖 原作者：K同学啊任务说明数据集中提供了火灾温度（Tem1）、一氧化碳浓度（CO 1）、烟雾浓度（Soot 1）随着时…

🍨 本文为🔗365天深度学习训练营中的学习记录博客
🍖 原作者：K同学啊

任务说明

数据集中提供了火灾温度（Tem1）、一氧化碳浓度（CO 1）、烟雾浓度（Soot 1）随着时间变化数据，我们需要根据这些数据对未来某一时刻的火灾温度做出预测（本次任务仅供学习）。

🍺要求：

了解LSTM是什么，并使用其构建一个完整的程序。
R²达到0.83。

🍻拔高：

使用第1-8个时刻的数据预测第9~10个时刻的温度数据。

🏡 我的环境：

● 语言环境：Python3.12.4
● 编译器：Jupyter Lab
● 深度学习框架：pyTorch
● 数据地址：🔗百度网盘

一、前期准备以及数据预处理

1、导入数据

和之前一样我们先将数据导入并使用pandas查看数据的构造。

import torch.nn.functional as F
import numpy as np
import pandas as pd
import torch
from torch import nnfile_path = 'D:/OneDrive/code learning(python and matlab and latex)/365camp/data/woodpine2.csv'
data = pd.read_csv(file_path)
data

代码输出：

         Time   Tem1      CO 1    Soot 1
0       0.000   25.0  0.000000  0.000000
1       0.228   25.0  0.000000  0.000000
2       0.456   25.0  0.000000  0.000000
3       0.685   25.0  0.000000  0.000000
4       0.913   25.0  0.000000  0.000000
...       ...    ...       ...       ...
5943  366.000  295.0  0.000077  0.000496
5944  366.000  294.0  0.000077  0.000494
5945  367.000  292.0  0.000077  0.000491
5946  367.000  291.0  0.000076  0.000489
5947  367.000  290.0  0.000076  0.000487[5948 rows x 4 columns]

2、数据集可视化

import matplotlib.pyplot as plt
import seaborn as snsplt.rcParams['savefig.dpi'] = 500
plt.rcParams['figure.dpi'] = 500fig, ax = plt.subplots(1, 3, constrained_layout=True, figsize=(14, 3))sns.lineplot(data = data['Tem1'], ax = ax[0])
sns.lineplot(data = data['CO 1'], ax = ax[1])
sns.lineplot(data = data['Soot 1'], ax = ax[2], color='r')
plt.show()

代码输出：
在这里插入图片描述

dataFrame = data.iloc[:, 1:]
dataFrame

代码输出：

       Tem1      CO 1    Soot 1
0      25.0  0.000000  0.000000
1      25.0  0.000000  0.000000
2      25.0  0.000000  0.000000
3      25.0  0.000000  0.000000
4      25.0  0.000000  0.000000
...     ...       ...       ...
5943  295.0  0.000077  0.000496
5944  294.0  0.000077  0.000494
5945  292.0  0.000077  0.000491
5946  291.0  0.000076  0.000489
5947  290.0  0.000076  0.000487[5948 rows x 3 columns]

二、构建数据集

1、数据集预处理

from sklearn.preprocessing import MinMaxScalerdataFrame = data.iloc[:, 1:].copy()
sc = MinMaxScaler(feature_range=(0, 1))for i in ['CO 1', 'Soot 1', 'Tem1']:dataFrame[i] = sc.fit_transform(dataFrame[i].values.reshape(-1, 1))dataFrame.shape

这段代码受限从sklearn库中导入了MinMaxScaler类，从data数据框中选择所有的行和第二列开始的所有的列进行创建一个新的数据框架dataFrame。copy()方法用于确保dataFrame是data的一个独立副本，随后创建了一个 MinMaxScaler 实例，设置归一化的范围为 0 到 1。MinMaxScaler 会将数据缩放到指定的范围内，通常用于将特征值标准化，以便于模型训练。然后这段代码对 dataFrame 中的三列（‘CO 1’、‘Soot 1’ 和 ‘Tem1’）进行归一化处理。fit_transform 方法会计算每列的最小值和最大值，并将其缩放到 0 到 1 的范围。reshape(-1, 1) 是为了将数据转换为二维数组的形式，因为 MinMaxScaler 需要这样的输入格式。
代码输出：

(5948, 3)

2、设置x和y

width_X = 8
width_y = 1##取前8个时间段的Tem1、CO 1、Soot 1为X，第9个时间段的Tem1为y。
X = []
y = []in_start = 0for _, _ in data.iterrows():in_end  = in_start + width_Xout_end = in_end   + width_yif out_end < len(dataFrame):X_ = np.array(dataFrame.iloc[in_start:in_end , ])y_ = np.array(dataFrame.iloc[in_end  :out_end, 0])X.append(X_)y.append(y_)in_start += 1X = np.array(X)
y = np.array(y).reshape(-1,width_y,1)X.shape, y.shape

代码输出：

((5938, 8, 3), (5938, 2, 1))

检查数据集中是否有空值：

print(np.any(np.isnan(X)))
print(np.any(np.isnan(y)))

代码输出：

False
False

X_train = torch.tensor(np.array(X[:5000]), dtype=torch.float32)
y_train = torch.tensor(np.array(y[:5000]), dtype=torch.float32)X_test  = torch.tensor(np.array(X[5000:]), dtype=torch.float32)
y_test  = torch.tensor(np.array(y[5000:]), dtype=torch.float32)
X_train.shape, y_train.shape

代码输出：

(torch.Size([5000, 8, 3]), torch.Size([5000, 2, 1]))

from torch.utils.data import TensorDataset, DataLoadertrain_dl = DataLoader(TensorDataset(X_train, y_train),batch_size=64, shuffle=True)test_dl  = DataLoader(TensorDataset(X_test, y_test),batch_size=64, shuffle=True)

三、模型训练

1、构建模型

class model_lstm1(nn.Module):def __init__(self):super(model_lstm1, self).__init__()self.lstm0 = nn.LSTM(input_size=3 ,hidden_size=320, num_layers=1, batch_first=True)self.lstm1 = nn.LSTM(input_size=320 ,hidden_size=320, num_layers=1, batch_first=True)self.fc0   = nn.Linear(320, 1)def forward(self, x):out, hidden1 = self.lstm0(x) out, _ = self.lstm1(out, hidden1) out    = self.fc0(out) return out[:, -1:, :]#取1个预测值,否则经过lstm会得到8*1个预测model1 = model_lstm1()
model1

代码输出：

model_lstm1((lstm0): LSTM(3, 320, batch_first=True)(lstm1): LSTM(320, 320, batch_first=True)(fc0): Linear(in_features=320, out_features=1, bias=True)
)

2、定义训练和测试函数

# 训练循环
import copy
def train(train_dl, model, loss_fn, opt, lr_scheduler=None):size        = len(train_dl.dataset)  num_batches = len(train_dl)   train_loss  = 0  # 初始化训练损失和正确率for x, y in train_dl:  x, y = x.to(device), y.to(device)# 计算预测误差pred = model(x)          # 网络输出loss = loss_fn(pred, y)  # 计算网络输出和真实值之间的差距# 反向传播opt.zero_grad()  # grad属性归零loss.backward()  # 反向传播opt.step()       # 每一步自动更新# 记录losstrain_loss += loss.item()if lr_scheduler is not None:lr_scheduler.step()print("learning rate = {:.5f}".format(opt.param_groups[0]['lr']), end="  ")train_loss /= num_batchesreturn train_lossdef test (dataloader, model, loss_fn):size        = len(dataloader.dataset)  # 测试集的大小num_batches = len(dataloader)          # 批次数目test_loss   = 0# 当不进行训练时，停止梯度更新，节省计算内存消耗with torch.no_grad():for x, y in dataloader:x, y = x.to(device), y.to(device)# 计算lossy_pred = model(x)loss        = loss_fn(y_pred, y)test_loss += loss.item()test_loss /= num_batchesreturn test_loss

3、训练函数

#%%
#设置GPU训练
device=torch.device("cuda" if torch.cuda.is_available() else "cpu")
device#训练模型
model = model_lstm1()
model = model.to(device)
loss_fn    = nn.MSELoss() # 创建损失函数
learn_rate = 1e-1   # 学习率
opt        = torch.optim.SGD(model.parameters(),lr=learn_rate,weight_decay=1e-4)
epochs     = 50
train_loss = []
test_loss  = []
lr_scheduler = torch.optim.lr_scheduler.CosineAnnealingLR(opt,epochs, last_epoch=-1) for epoch in range(epochs):model.train()epoch_train_loss = train(train_dl, model, loss_fn, opt, lr_scheduler)model.eval()epoch_test_loss = test(test_dl, model, loss_fn)train_loss.append(epoch_train_loss)test_loss.append(epoch_test_loss)template = ('Epoch:{:2d}, Train_loss:{:.5f}, Test_loss:{:.5f}')print(template.format(epoch+1, epoch_train_loss,  epoch_test_loss))print("="*20, 'Done', "="*20)

代码输出：

learning rate = 0.09990  Epoch: 1, Train_loss:0.05296, Test_loss:0.20052
learning rate = 0.09961  Epoch: 2, Train_loss:0.04132, Test_loss:0.14725
learning rate = 0.09911  Epoch: 3, Train_loss:0.02477, Test_loss:0.06766
learning rate = 0.09843  Epoch: 4, Train_loss:0.00568, Test_loss:0.00505
learning rate = 0.09755  Epoch: 5, Train_loss:0.00037, Test_loss:0.00088
learning rate = 0.09649  Epoch: 6, Train_loss:0.00016, Test_loss:0.00079
learning rate = 0.09524  Epoch: 7, Train_loss:0.00015, Test_loss:0.00069
learning rate = 0.09382  Epoch: 8, Train_loss:0.00015, Test_loss:0.00060
learning rate = 0.09222  Epoch: 9, Train_loss:0.00015, Test_loss:0.00072
learning rate = 0.09045  Epoch:10, Train_loss:0.00015, Test_loss:0.00062
learning rate = 0.08853  Epoch:11, Train_loss:0.00015, Test_loss:0.00083
learning rate = 0.08645  Epoch:12, Train_loss:0.00016, Test_loss:0.00065
learning rate = 0.08423  Epoch:13, Train_loss:0.00015, Test_loss:0.00059
learning rate = 0.08187  Epoch:14, Train_loss:0.00015, Test_loss:0.00060
learning rate = 0.07939  Epoch:15, Train_loss:0.00015, Test_loss:0.00064
learning rate = 0.07679  Epoch:16, Train_loss:0.00015, Test_loss:0.00068
learning rate = 0.07409  Epoch:17, Train_loss:0.00015, Test_loss:0.00067
learning rate = 0.07129  Epoch:18, Train_loss:0.00015, Test_loss:0.00075
learning rate = 0.06841  Epoch:19, Train_loss:0.00015, Test_loss:0.00066
learning rate = 0.06545  Epoch:20, Train_loss:0.00015, Test_loss:0.00062
learning rate = 0.06243  Epoch:21, Train_loss:0.00015, Test_loss:0.00069
learning rate = 0.05937  Epoch:22, Train_loss:0.00015, Test_loss:0.00072
learning rate = 0.05627  Epoch:23, Train_loss:0.00015, Test_loss:0.00064
learning rate = 0.05314  Epoch:24, Train_loss:0.00015, Test_loss:0.00070
learning rate = 0.05000  Epoch:25, Train_loss:0.00015, Test_loss:0.00067
learning rate = 0.04686  Epoch:26, Train_loss:0.00015, Test_loss:0.00078
learning rate = 0.04373  Epoch:27, Train_loss:0.00015, Test_loss:0.00067
learning rate = 0.04063  Epoch:28, Train_loss:0.00015, Test_loss:0.00065
learning rate = 0.03757  Epoch:29, Train_loss:0.00015, Test_loss:0.00074
learning rate = 0.03455  Epoch:30, Train_loss:0.00015, Test_loss:0.00068
learning rate = 0.03159  Epoch:31, Train_loss:0.00015, Test_loss:0.00065
learning rate = 0.02871  Epoch:32, Train_loss:0.00015, Test_loss:0.00064
learning rate = 0.02591  Epoch:33, Train_loss:0.00015, Test_loss:0.00066
learning rate = 0.02321  Epoch:34, Train_loss:0.00015, Test_loss:0.00063
learning rate = 0.02061  Epoch:35, Train_loss:0.00015, Test_loss:0.00066
learning rate = 0.01813  Epoch:36, Train_loss:0.00015, Test_loss:0.00065
learning rate = 0.01577  Epoch:37, Train_loss:0.00015, Test_loss:0.00068
learning rate = 0.01355  Epoch:38, Train_loss:0.00015, Test_loss:0.00067
learning rate = 0.01147  Epoch:39, Train_loss:0.00015, Test_loss:0.00067
learning rate = 0.00955  Epoch:40, Train_loss:0.00015, Test_loss:0.00064
learning rate = 0.00778  Epoch:41, Train_loss:0.00014, Test_loss:0.00065
learning rate = 0.00618  Epoch:42, Train_loss:0.00015, Test_loss:0.00064
learning rate = 0.00476  Epoch:43, Train_loss:0.00015, Test_loss:0.00065
learning rate = 0.00351  Epoch:44, Train_loss:0.00015, Test_loss:0.00065
learning rate = 0.00245  Epoch:45, Train_loss:0.00015, Test_loss:0.00066
learning rate = 0.00157  Epoch:46, Train_loss:0.00015, Test_loss:0.00065
learning rate = 0.00089  Epoch:47, Train_loss:0.00015, Test_loss:0.00066
learning rate = 0.00039  Epoch:48, Train_loss:0.00015, Test_loss:0.00066
learning rate = 0.00010  Epoch:49, Train_loss:0.00015, Test_loss:0.00066
learning rate = 0.00000  Epoch:50, Train_loss:0.00015, Test_loss:0.00066
==================== Done ====================

四、模型评估

1、Loss图

import matplotlib.pyplot as plt
from datetime import datetime
current_time = datetime.now() # 获取当前时间plt.figure(figsize=(5, 3),dpi=120)plt.plot(train_loss    , label='LSTM Training Loss')
plt.plot(test_loss, label='LSTM Validation Loss')plt.title('Training and Validation Loss')
plt.xlabel(current_time) # 打卡请带上时间戳，否则代码截图无效
plt.legend()
plt.show()

代码输出：
在这里插入图片描述

2、调用模型预测：

model = model.to(device)  # 将模型移动到 GPU
X_test = X_test.to(device)  # 将测试数据移动到 GPU# 测试集输入模型进行预测，然后转换回 CPU 进行反归一化
predicted_y_lstm = sc.inverse_transform(model(X_test).detach().cpu().numpy().reshape(-1, 1))
y_test_1 = sc.inverse_transform(y_test.reshape(-1, 1))
y_test_one = [i[0] for i in y_test_1]
predicted_y_lstm_one = [i[0] for i in predicted_y_lstm]plt.figure(figsize=(5, 3), dpi=120)
# 画出真实数据和预测数据的对比曲线
plt.plot(y_test_one[:2000], color='red', label='real_temp')
plt.plot(predicted_y_lstm_one[:2000], color='blue', label='prediction')plt.title('Title')
plt.xlabel('X')
plt.ylabel('Y')
plt.legend()
plt.show()

代码输出：
在这里插入图片描述

3、R²评估

from sklearn import metrics
"""
RMSE ：均方根误差  ----->  对均方误差开方
R2   ：决定系数，可以简单理解为反映模型拟合优度的重要的统计量
"""
RMSE_lstm  = metrics.mean_squared_error(predicted_y_lstm_one, y_test_1)**0.5
R2_lstm    = metrics.r2_score(predicted_y_lstm_one, y_test_1)print('均方根误差: %.5f' % RMSE_lstm)
print('R2: %.5f' % R2_lstm)
``
代码输出：```python
均方根误差: 7.21383
R2: 0.81756

五、提升：使用第1-8个时刻的数据预测第9~10个时刻的温度数据

模型改造：

class model_lstm2(nn.Module):def __init__(self):super(model_lstm2, self).__init__()self.lstm0 = nn.LSTM(input_size=3, hidden_size=320, num_layers=1, batch_first=True)self.lstm1 = nn.LSTM(input_size=320, hidden_size=320, num_layers=1, batch_first=True)# 修改 fc0，输出 2 个值self.fc0   = nn.Linear(320, 2)def forward(self, x):out, hidden1 = self.lstm0(x) out, _ = self.lstm1(out, hidden1) # 取最后一个时间步的隐藏状态，经过 fc0 得到 (batch, 2)out = self.fc0(out[:, -1, :])# 重新调整形状为 (batch, 2, 1)return out.unsqueeze(2)model2 = model_lstm2()
model2

使输出后两个温度，重新训练模型：

learning rate = 0.09990  Epoch: 1, Train_loss:0.05822, Test_loss:0.19761
learning rate = 0.09961  Epoch: 2, Train_loss:0.04638, Test_loss:0.17118
learning rate = 0.09911  Epoch: 3, Train_loss:0.03971, Test_loss:0.13120
learning rate = 0.09843  Epoch: 4, Train_loss:0.02665, Test_loss:0.07101
learning rate = 0.09755  Epoch: 5, Train_loss:0.00976, Test_loss:0.01794
learning rate = 0.09649  Epoch: 6, Train_loss:0.00118, Test_loss:0.00188
learning rate = 0.09524  Epoch: 7, Train_loss:0.00019, Test_loss:0.00082
learning rate = 0.09382  Epoch: 8, Train_loss:0.00015, Test_loss:0.00079
learning rate = 0.09222  Epoch: 9, Train_loss:0.00015, Test_loss:0.00073
learning rate = 0.09045  Epoch:10, Train_loss:0.00015, Test_loss:0.00093
learning rate = 0.08853  Epoch:11, Train_loss:0.00015, Test_loss:0.00075
learning rate = 0.08645  Epoch:12, Train_loss:0.00015, Test_loss:0.00077
learning rate = 0.08423  Epoch:13, Train_loss:0.00015, Test_loss:0.00078
learning rate = 0.08187  Epoch:14, Train_loss:0.00015, Test_loss:0.00079
learning rate = 0.07939  Epoch:15, Train_loss:0.00014, Test_loss:0.00074
learning rate = 0.07679  Epoch:16, Train_loss:0.00014, Test_loss:0.00077
learning rate = 0.07409  Epoch:17, Train_loss:0.00014, Test_loss:0.00074
learning rate = 0.07129  Epoch:18, Train_loss:0.00014, Test_loss:0.00080
learning rate = 0.06841  Epoch:19, Train_loss:0.00014, Test_loss:0.00080
learning rate = 0.06545  Epoch:20, Train_loss:0.00014, Test_loss:0.00080
learning rate = 0.06243  Epoch:21, Train_loss:0.00014, Test_loss:0.00071
learning rate = 0.05937  Epoch:22, Train_loss:0.00015, Test_loss:0.00073
learning rate = 0.05627  Epoch:23, Train_loss:0.00014, Test_loss:0.00075
learning rate = 0.05314  Epoch:24, Train_loss:0.00014, Test_loss:0.00069
learning rate = 0.05000  Epoch:25, Train_loss:0.00014, Test_loss:0.00071
learning rate = 0.04686  Epoch:26, Train_loss:0.00014, Test_loss:0.00074
learning rate = 0.04373  Epoch:27, Train_loss:0.00014, Test_loss:0.00079
learning rate = 0.04063  Epoch:28, Train_loss:0.00014, Test_loss:0.00076
learning rate = 0.03757  Epoch:29, Train_loss:0.00014, Test_loss:0.00079
learning rate = 0.03455  Epoch:30, Train_loss:0.00014, Test_loss:0.00072
learning rate = 0.03159  Epoch:31, Train_loss:0.00014, Test_loss:0.00080
learning rate = 0.02871  Epoch:32, Train_loss:0.00014, Test_loss:0.00075
learning rate = 0.02591  Epoch:33, Train_loss:0.00014, Test_loss:0.00074
learning rate = 0.02321  Epoch:34, Train_loss:0.00015, Test_loss:0.00076
learning rate = 0.02061  Epoch:35, Train_loss:0.00014, Test_loss:0.00075
learning rate = 0.01813  Epoch:36, Train_loss:0.00014, Test_loss:0.00076
learning rate = 0.01577  Epoch:37, Train_loss:0.00014, Test_loss:0.00075
learning rate = 0.01355  Epoch:38, Train_loss:0.00014, Test_loss:0.00077
learning rate = 0.01147  Epoch:39, Train_loss:0.00014, Test_loss:0.00073
learning rate = 0.00955  Epoch:40, Train_loss:0.00014, Test_loss:0.00077
learning rate = 0.00778  Epoch:41, Train_loss:0.00014, Test_loss:0.00075
learning rate = 0.00618  Epoch:42, Train_loss:0.00014, Test_loss:0.00077
learning rate = 0.00476  Epoch:43, Train_loss:0.00014, Test_loss:0.00077
learning rate = 0.00351  Epoch:44, Train_loss:0.00014, Test_loss:0.00076
learning rate = 0.00245  Epoch:45, Train_loss:0.00014, Test_loss:0.00076
learning rate = 0.00157  Epoch:46, Train_loss:0.00014, Test_loss:0.00076
learning rate = 0.00089  Epoch:47, Train_loss:0.00014, Test_loss:0.00076
learning rate = 0.00039  Epoch:48, Train_loss:0.00014, Test_loss:0.00077
learning rate = 0.00010  Epoch:49, Train_loss:0.00014, Test_loss:0.00076
learning rate = 0.00000  Epoch:50, Train_loss:0.00014, Test_loss:0.00076
==================== Done ====================