当前位置：首页 > news >正文

第R6周：LSTM实现糖尿病探索与预测

news 来源：原创 2025/6/10 12:17:23

🍨 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

文章目录

1、设置GPU
2、导入数据
3、数据检查
4、数据分布分析
5、数据集构建
6、定义模型
7、定义训练函数
8、定义测试函数
9、训练模型
10、结果可视化

电脑环境：
语言环境：Python 3.8.0
深度学习：torch 2.5.1+cu124

1、设置GPU

import torch.nn.functional as F
import torch.nn as nn
import torch, torchvision

device = torch.device("cuda" if torch.cuda.is_available() else "cpu")
device

2、导入数据

import numpy as np
import pandas as pd
import seaborn as sns
from sklearn.model_selection import train_test_split
import matplotlib.pyplot as plt
# 支持中文
plt.rcParams['savefig.dpi'] = 500
plt.rcParams['figure.dpi'] = 500
plt.rcParams['font.sans-serif'] = ['SimHei']  # 用来正常显示中文标签

import warnings
warnings.filterwarnings("ignore")

DataFrame = pd.read_excel('./dia.xls')
DataFrame.head()

在这里插入图片描述

DataFrame.shape

代码输出：

(1006, 16)

3、数据检查

# 查看是否有缺失值
print("数据缺失值------------------")
print(DataFrame.isnull().sum())

代码输出：

数据缺失值------------------
卡号 0
性别 0
年龄 0
高密度脂蛋白胆固醇 0
低密度脂蛋白胆固醇 0
极低密度脂蛋白胆固醇 0
甘油三酯 0
总胆固醇 0
脉搏 0
舒张压 0
高血压史 0
尿素氮 0
尿酸 0
肌酐 0
体重检查结果 0
是否糖尿病 0
dtype: int64

# 查看数据是否有重复值
print("数据重复值------------------")
print('数据的重复值为：'f'{DataFrame.duplicated().sum()}')

代码输出：

数据重复值------------------
数据的重复值为：0

4、数据分布分析

feature_map = {
            '年龄': '年龄',
            '高密度脂蛋白胆固醇': '高密度脂蛋白胆固醇',
            '低密度脂蛋白胆固醇': '低密度脂蛋白胆固醇',
            '极低密度脂蛋白胆固醇': '极低密度脂蛋白胆固醇',
            '甘油三酯': '甘油三酯',
            '总胆固醇': '总胆固醇',
            '脉搏': '脉搏',
            '舒张压': '舒张压',
            '高血压史': '高血压史',
            '尿素氮': '尿素氮',
            '尿酸': '尿酸',
            '肌酐': '肌酐',
            '体重检查结果': '体重检查结果'
}

plt.figure(figsize=(15, 10))
for i, (col, col_name) in enumerate(feature_map.items(), 1):
    plt.subplot(3, 5, i)
    sns.boxplot(x=DataFrame['是否糖尿病'], y=DataFrame[col])
    plt.title(f'{col_name}的箱线图', fontsize=14)
    plt.ylabel('数值', fontsize=12)
    plt.grid(axis='y', linestyle='--', alpha=0.7)
plt.tight_layout()
plt.show()

在这里插入图片描述

5、数据集构建

from sklearn.preprocessing import StandardScaler

# '高密度脂蛋白胆固醇'字段与糖尿病负相关，故在X 中去掉该字段
X = DataFrame.drop(['卡号', '是否糖尿病', '高密度脂蛋白胆固醇'], axis=1)
y = DataFrame['是否糖尿病']

# sc_X = StandardScaler
# X = sc_X.fit_transform(X)

X = torch.tensor(np.array(X), dtype=torch.float32)
y = torch.tensor(np.array(y), dtype=torch.int64)

train_X, test_X, train_y, test_y = train_test_split(X, y, test_size=0.2, random_state=1)

train_X.shape, train_y.shape

代码输出：

(torch.Size([804, 13]), torch.Size([804]))

from torch.utils.data import TensorDataset, DataLoader

train_dl = DataLoader(TensorDataset(train_X, train_y), 
                      batch_size=64, 
                      shuffle=False)
test_dl = DataLoader(TensorDataset(test_X, test_y), 
                      batch_size=64, 
                      shuffle=False)

6、定义模型

class model_lstm(nn.Module):
    def __init__(self):
        super(model_lstm, self).__init__()
        self.lstm0 = nn.LSTM(input_size=13, hidden_size=300, num_layers=1, batch_first=True)
        self.lstm1 = nn.LSTM(input_size=300, hidden_size=300, num_layers=1, batch_first=True)
        self.lstm2 = nn.LSTM(input_size=300, hidden_size=300, num_layers=1, batch_first=True)
        self.dropout = nn.Dropout(0.3)
        self.fc0 = nn.Linear(300, 2)

    def forward(self, x):
        out, hidden1 = self.lstm0(x)
        out, hidden2 = self.lstm1(out, hidden1)
        out, _ = self.lstm2(out, hidden2)
        out = self.dropout(out)
        out = self.fc0(out)
        return out

model = model_lstm().to(device)
model

代码输出：

model_lstm(
(lstm0): LSTM(13, 300, batch_first=True)
(lstm1): LSTM(300, 300, batch_first=True)
(lstm2): LSTM(300, 300, batch_first=True)
(dropout): Dropout(p=0.3, inplace=False)
(fc0): Linear(in_features=300, out_features=2, bias=True)
)

7、定义训练函数

# 训练循环
def train(dataloader, model, loss_fn, optimizer):
    size = len(dataloader.dataset)  # 训练集的大小
    num_batches = len(dataloader)   # 批次数目, (size/batch_size，向上取整)

    train_loss, train_acc = 0, 0  # 初始化训练损失和正确率

    for X, y in dataloader:  # 获取图片及其标签
        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，targets为真实值，计算二者差值即为损失

        # 反向传播
        optimizer.zero_grad()  # grad属性归零
        loss.backward()        # 反向传播
        optimizer.step()       # 每一步自动更新

        # 记录acc与loss
        train_acc  += (pred.argmax(1) == y).type(torch.float).sum().item()
        train_loss += loss.item()

    train_acc  /= size
    train_loss /= num_batches

    return train_acc, train_loss

8、定义测试函数

def test (dataloader, model, loss_fn):
    size        = len(dataloader.dataset)  # 测试集的大小
    num_batches = len(dataloader)          # 批次数目, (size/batch_size，向上取整)
    test_loss, test_acc = 0, 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 imgs, target in dataloader:
            imgs, target = imgs.to(device), target.to(device)

            # 计算loss
            target_pred = model(imgs)
            loss        = loss_fn(target_pred, target)

            test_loss += loss.item()
            test_acc  += (target_pred.argmax(1) == target).type(torch.float).sum().item()

    test_acc  /= size
    test_loss /= num_batches

    return test_acc, test_loss

9、训练模型

loss_fn    = nn.CrossEntropyLoss() # 创建损失函数
learn_rate = 1e-3
opt = torch.optim.Adam(model.parameters(), lr= learn_rate)

epochs     = 30

train_loss = []
train_acc  = []
test_loss  = []
test_acc   = []

for epoch in range(epochs):

    model.train()
    epoch_train_acc, epoch_train_loss = train(train_dl, model, loss_fn, opt)

    model.eval()
    epoch_test_acc, epoch_test_loss = test(test_dl, model, loss_fn)

    train_acc.append(epoch_train_acc)
    train_loss.append(epoch_train_loss)
    test_acc.append(epoch_test_acc)
    test_loss.append(epoch_test_loss)

    # 获取当前的学习率
    lr = opt.state_dict()['param_groups'][0]['lr']

    template = ('Epoch:{:2d}, Train_acc:{:.1f}%, Train_loss:{:.3f}, Test_acc:{:.1f}%, Test_loss:{:.3f}, Lr:{:.2E}')
    print(template.format(epoch+1, epoch_train_acc*100, epoch_train_loss,
                          epoch_test_acc*100, epoch_test_loss, lr))

print('Done')

代码输出：

Epoch: 1, Train_acc:53.0%, Train_loss:0.695, Test_acc:53.0%, Test_loss:0.716, Lr:1.00E-03
Epoch: 2, Train_acc:56.0%, Train_loss:0.686, Test_acc:53.0%, Test_loss:0.718, Lr:1.00E-03
Epoch: 3, Train_acc:56.2%, Train_loss:0.683, Test_acc:53.0%, Test_loss:0.725, Lr:1.00E-03
Epoch: 4, Train_acc:56.6%, Train_loss:0.671, Test_acc:57.4%, Test_loss:0.701, Lr:1.00E-03
Epoch: 5, Train_acc:63.7%, Train_loss:0.621, Test_acc:62.4%, Test_loss:0.693, Lr:1.00E-03
Epoch: 6, Train_acc:69.2%, Train_loss:0.554, Test_acc:62.9%, Test_loss:0.598, Lr:1.00E-03
Epoch: 7, Train_acc:71.9%, Train_loss:0.509, Test_acc:66.3%, Test_loss:0.605, Lr:1.00E-03
Epoch: 8, Train_acc:74.4%, Train_loss:0.491, Test_acc:67.8%, Test_loss:0.614, Lr:1.00E-03
Epoch: 9, Train_acc:77.5%, Train_loss:0.452, Test_acc:70.8%, Test_loss:0.633, Lr:1.00E-03
Epoch:10, Train_acc:79.0%, Train_loss:0.430, Test_acc:73.8%, Test_loss:0.584, Lr:1.00E-03
Epoch:11, Train_acc:79.4%, Train_loss:0.416, Test_acc:75.7%, Test_loss:0.681, Lr:1.00E-03
Epoch:12, Train_acc:79.6%, Train_loss:0.414, Test_acc:70.3%, Test_loss:0.792, Lr:1.00E-03
Epoch:13, Train_acc:83.5%, Train_loss:0.376, Test_acc:71.8%, Test_loss:0.787, Lr:1.00E-03
Epoch:14, Train_acc:83.3%, Train_loss:0.354, Test_acc:72.3%, Test_loss:0.817, Lr:1.00E-03
Epoch:15, Train_acc:84.0%, Train_loss:0.338, Test_acc:69.3%, Test_loss:0.811, Lr:1.00E-03
Epoch:16, Train_acc:85.2%, Train_loss:0.321, Test_acc:70.3%, Test_loss:0.809, Lr:1.00E-03
Epoch:17, Train_acc:85.1%, Train_loss:0.320, Test_acc:67.8%, Test_loss:1.088, Lr:1.00E-03
Epoch:18, Train_acc:83.1%, Train_loss:0.345, Test_acc:67.8%, Test_loss:0.774, Lr:1.00E-03
Epoch:19, Train_acc:86.1%, Train_loss:0.304, Test_acc:68.8%, Test_loss:0.865, Lr:1.00E-03
Epoch:20, Train_acc:84.7%, Train_loss:0.313, Test_acc:72.3%, Test_loss:0.908, Lr:1.00E-03
Epoch:21, Train_acc:87.3%, Train_loss:0.284, Test_acc:72.8%, Test_loss:0.939, Lr:1.00E-03
Epoch:22, Train_acc:88.8%, Train_loss:0.252, Test_acc:71.8%, Test_loss:1.224, Lr:1.00E-03
Epoch:23, Train_acc:90.5%, Train_loss:0.222, Test_acc:66.8%, Test_loss:1.156, Lr:1.00E-03
Epoch:24, Train_acc:89.8%, Train_loss:0.206, Test_acc:67.8%, Test_loss:1.117, Lr:1.00E-03
Epoch:25, Train_acc:93.0%, Train_loss:0.175, Test_acc:70.8%, Test_loss:1.227, Lr:1.00E-03
Epoch:26, Train_acc:92.9%, Train_loss:0.170, Test_acc:67.8%, Test_loss:1.311, Lr:1.00E-03
Epoch:27, Train_acc:90.9%, Train_loss:0.184, Test_acc:66.3%, Test_loss:1.432, Lr:1.00E-03
Epoch:28, Train_acc:92.0%, Train_loss:0.180, Test_acc:68.3%, Test_loss:1.397, Lr:1.00E-03
Epoch:29, Train_acc:92.0%, Train_loss:0.191, Test_acc:65.3%, Test_loss:1.339, Lr:1.00E-03
Epoch:30, Train_acc:90.9%, Train_loss:0.208, Test_acc:67.3%, Test_loss:1.235, Lr:1.00E-03
Done

10、结果可视化

import matplotlib.pyplot as plt
#隐藏警告
import warnings
warnings.filterwarnings("ignore")               #忽略警告信息
plt.rcParams['font.sans-serif']    = ['SimHei'] # 用来正常显示中文标签
plt.rcParams['axes.unicode_minus'] = False      # 用来正常显示负号
plt.rcParams['figure.dpi']         = 100        #分辨率

from datetime import datetime
current_time = datetime.now()

epochs_range = range(epochs)

plt.figure(figsize=(12, 3))
plt.subplot(1, 2, 1)

plt.plot(epochs_range, train_acc, label='Training Accuracy')
plt.plot(epochs_range, test_acc, label='Test Accuracy')
plt.legend(loc='lower right')
plt.title('Training and Validation Accuracy')
plt.xlabel(current_time)

plt.subplot(1, 2, 2)
plt.plot(epochs_range, train_loss, label='Training Loss')
plt.plot(epochs_range, test_loss, label='Test Loss')
plt.legend(loc='upper right')
plt.title('Training and Validation Loss')
plt.show()