- 🍨 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
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 numpy as np
import pandas as pd
import matplotlib.pyplot as plt
import seaborn as sns
import torch
import torch.nn as nn
import torch.nn.functional as F
# 设置GPU
device = torch.device("cuda" if torch.cuda.is_available() else "cpu")
print(device)
2、导入数据
df = pd.read_csv('/alzheimers_disease_data.csv')
# 删除最后一列和第一列
df = df.iloc[:, 1:-1]
df.head()
Age Gender Ethnicity EducationLevel BMI Smoking AlcoholConsumption PhysicalActivity DietQuality SleepQuality ... FunctionalAssessment MemoryComplaints BehavioralProblems ADL Confusion Disorientation PersonalityChanges DifficultyCompletingTasks Forgetfulness Diagnosis
0 73 0 0 2 22.927749 0 13.297218 6.327112 1.347214 9.025679 ... 6.518877 0 0 1.725883 0 0 0 1 0 0
1 89 0 0 0 26.827681 0 4.542524 7.619885 0.518767 7.151293 ... 7.118696 0 0 2.592424 0 0 0 0 1 0
2 73 0 3 1 17.795882 0 19.555085 7.844988 1.826335 9.673574 ... 5.895077 0 0 7.119548 0 1 0 1 0 0
3 74 1 0 1 33.800817 1 12.209266 8.428001 7.435604 8.392554 ... 8.965106 0 1 6.481226 0 0 0 0 0 0
4 89 0 0 0 20.716974 0 18.454356 6.310461 0.795498 5.597238 ... 6.045039 0 0 0.014691 0 0 1 1 0 0
5 rows × 33 columns
3、数据标准化
from sklearn.model_selection import train_test_split
from sklearn.preprocessing import StandardScaler
X = df.iloc[:, :-1]
y = df.iloc[:, -1]
# 将每一列特征标准化为标准正态分布，注意，标准化是针对每一列而言的
scaler = StandardScaler()
X = scaler.fit_transform(X)
4、划分数据集
X = torch.tensor(np.array(X), dtype=torch.float32)
y = torch.tensor(np.array(y), dtype=torch.int64)
X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.1, random_state=1)
X_train.shape, y_train.shape
5、构建数据加载器
from torch.utils.data import TensorDataset, DataLoader
train_dl = DataLoader(TensorDataset(X_train, y_train),
batch_size=32,
shuffle=False)
test_dl = DataLoader(TensorDataset(X_test, y_test),
batch_size=32,
shuffle=False)
6、模型构建
class model_rnn(nn.Module):
def __init__(self):
super(model_rnn, self).__init__()
self.rnn0 = nn.RNN(input_size=32, hidden_size=200, num_layers=1, batch_first=True)
self.fc0 = nn.Linear(200, 50)
self.fc1 = nn.Linear(50, 2)
def forward(self, x):
out, hidden1 = self.rnn0(x)
out = self.fc0(out)
out = self.fc1(out)
return out
model = model_rnn().to(device)
model
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、正式训练
loss_fn = nn.CrossEntropyLoss() # 创建损失函数
learn_rate = 5e-5
opt = torch.optim.Adam(model.parameters(), lr= learn_rate)
epochs = 50
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')
Epoch: 1, Train_acc:61.4%, Train_loss:0.672, Test_acc:65.1%, Test_loss:0.656, Lr:5.00E-05
Epoch: 2, Train_acc:67.3%, Train_loss:0.629, Test_acc:65.6%, Test_loss:0.622, Lr:5.00E-05
Epoch: 3, Train_acc:67.1%, Train_loss:0.592, Test_acc:66.5%, Test_loss:0.597, Lr:5.00E-05
Epoch: 4, Train_acc:69.5%, Train_loss:0.560, Test_acc:67.4%, Test_loss:0.571, Lr:5.00E-05
Epoch: 5, Train_acc:74.1%, Train_loss:0.528, Test_acc:70.7%, Test_loss:0.544, Lr:5.00E-05
Epoch: 6, Train_acc:78.5%, Train_loss:0.494, Test_acc:74.9%, Test_loss:0.517, Lr:5.00E-05
Epoch: 7, Train_acc:81.6%, Train_loss:0.463, Test_acc:76.3%, Test_loss:0.493, Lr:5.00E-05
Epoch: 8, Train_acc:83.1%, Train_loss:0.435, Test_acc:76.3%, Test_loss:0.472, Lr:5.00E-05
Epoch: 9, Train_acc:83.9%, Train_loss:0.412, Test_acc:77.7%, Test_loss:0.455, Lr:5.00E-05
Epoch:10, Train_acc:84.8%, Train_loss:0.394, Test_acc:80.0%, Test_loss:0.442, Lr:5.00E-05
Epoch:11, Train_acc:85.1%, Train_loss:0.380, Test_acc:80.9%, Test_loss:0.432, Lr:5.00E-05
Epoch:12, Train_acc:85.3%, Train_loss:0.368, Test_acc:80.5%, Test_loss:0.425, Lr:5.00E-05
Epoch:13, Train_acc:85.4%, Train_loss:0.360, Test_acc:80.0%, Test_loss:0.420, Lr:5.00E-05
Epoch:14, Train_acc:85.7%, Train_loss:0.353, Test_acc:79.5%, Test_loss:0.417, Lr:5.00E-05
Epoch:15, Train_acc:86.1%, Train_loss:0.347, Test_acc:79.5%, Test_loss:0.415, Lr:5.00E-05
Epoch:16, Train_acc:86.4%, Train_loss:0.342, Test_acc:79.5%, Test_loss:0.414, Lr:5.00E-05
Epoch:17, Train_acc:86.8%, Train_loss:0.338, Test_acc:79.1%, Test_loss:0.414, Lr:5.00E-05
Epoch:18, Train_acc:87.0%, Train_loss:0.335, Test_acc:78.1%, Test_loss:0.415, Lr:5.00E-05
Epoch:19, Train_acc:87.0%, Train_loss:0.331, Test_acc:77.2%, Test_loss:0.416, Lr:5.00E-05
Epoch:20, Train_acc:87.1%, Train_loss:0.329, Test_acc:76.7%, Test_loss:0.418, Lr:5.00E-05
Epoch:21, Train_acc:87.1%, Train_loss:0.326, Test_acc:76.7%, Test_loss:0.420, Lr:5.00E-05
Epoch:22, Train_acc:87.2%, Train_loss:0.323, Test_acc:77.7%, Test_loss:0.423, Lr:5.00E-05
Epoch:23, Train_acc:87.0%, Train_loss:0.321, Test_acc:77.7%, Test_loss:0.425, Lr:5.00E-05
Epoch:24, Train_acc:87.3%, Train_loss:0.319, Test_acc:77.7%, Test_loss:0.428, Lr:5.00E-05
Epoch:25, Train_acc:87.2%, Train_loss:0.317, Test_acc:78.1%, Test_loss:0.431, Lr:5.00E-05
Epoch:26, Train_acc:87.3%, Train_loss:0.315, Test_acc:77.7%, Test_loss:0.433, Lr:5.00E-05
Epoch:27, Train_acc:87.4%, Train_loss:0.313, Test_acc:78.1%, Test_loss:0.436, Lr:5.00E-05
Epoch:28, Train_acc:87.5%, Train_loss:0.311, Test_acc:77.7%, Test_loss:0.439, Lr:5.00E-05
Epoch:29, Train_acc:87.5%, Train_loss:0.310, Test_acc:78.1%, Test_loss:0.441, Lr:5.00E-05
Epoch:30, Train_acc:87.7%, Train_loss:0.308, Test_acc:78.1%, Test_loss:0.444, Lr:5.00E-05
Epoch:31, Train_acc:87.7%, Train_loss:0.306, Test_acc:78.6%, Test_loss:0.447, Lr:5.00E-05
Epoch:32, Train_acc:87.9%, Train_loss:0.305, Test_acc:76.7%, Test_loss:0.449, Lr:5.00E-05
Epoch:33, Train_acc:88.0%, Train_loss:0.303, Test_acc:76.3%, Test_loss:0.452, Lr:5.00E-05
Epoch:34, Train_acc:88.0%, Train_loss:0.301, Test_acc:75.8%, Test_loss:0.454, Lr:5.00E-05
Epoch:35, Train_acc:88.1%, Train_loss:0.300, Test_acc:76.3%, Test_loss:0.457, Lr:5.00E-05
Epoch:36, Train_acc:88.2%, Train_loss:0.298, Test_acc:76.3%, Test_loss:0.460, Lr:5.00E-05
Epoch:37, Train_acc:88.4%, Train_loss:0.296, Test_acc:77.2%, Test_loss:0.462, Lr:5.00E-05
Epoch:38, Train_acc:88.6%, Train_loss:0.294, Test_acc:77.2%, Test_loss:0.465, Lr:5.00E-05
Epoch:39, Train_acc:88.7%, Train_loss:0.293, Test_acc:77.2%, Test_loss:0.468, Lr:5.00E-05
Epoch:40, Train_acc:88.8%, Train_loss:0.291, Test_acc:77.2%, Test_loss:0.471, Lr:5.00E-05
Epoch:41, Train_acc:88.8%, Train_loss:0.289, Test_acc:77.2%, Test_loss:0.475, Lr:5.00E-05
Epoch:42, Train_acc:89.0%, Train_loss:0.287, Test_acc:77.2%, Test_loss:0.478, Lr:5.00E-05
Epoch:43, Train_acc:88.9%, Train_loss:0.285, Test_acc:77.7%, Test_loss:0.481, Lr:5.00E-05
Epoch:44, Train_acc:88.9%, Train_loss:0.283, Test_acc:77.7%, Test_loss:0.485, Lr:5.00E-05
Epoch:45, Train_acc:89.0%, Train_loss:0.282, Test_acc:78.1%, Test_loss:0.489, Lr:5.00E-05
Epoch:46, Train_acc:89.0%, Train_loss:0.280, Test_acc:78.1%, Test_loss:0.493, Lr:5.00E-05
Epoch:47, Train_acc:89.2%, Train_loss:0.278, Test_acc:78.1%, Test_loss:0.497, Lr:5.00E-05
Epoch:48, Train_acc:89.4%, Train_loss:0.276, Test_acc:78.1%, Test_loss:0.501, Lr:5.00E-05
Epoch:49, Train_acc:89.6%, Train_loss:0.274, Test_acc:78.6%, Test_loss:0.506, Lr:5.00E-05
Epoch:50, Train_acc:89.6%, Train_loss:0.272, Test_acc:78.1%, Test_loss:0.510, Lr:5.00E-05
Done
10、Loss与Accuracy图
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()
11、混淆矩阵
print("====================输入数据Shape为====================")
print("X_test.shape: ",X_test.shape)
print("y_test.shape: ",y_test.shape)
pred = model(X_test.to(device)).argmax(1).cpu().numpy()
print("====================输出数据Shape为====================")
print("pred.shape: ",pred.shape)
输入数据Shape为
X_test.shape: torch.Size([215, 32])
y_test.shape: torch.Size([215])
输出数据Shape为
pred.shape: (215,)
import matplotlib.pyplot as plt
from sklearn.metrics import confusion_matrix, ConfusionMatrixDisplay
# 计算混淆矩阵
cm = confusion_matrix(y_test, pred)
plt.figure(figsize=(6,5))
# plt.suptitle('Confusion Matrix')
sns.heatmap(cm, annot=True, fmt='d', cmap='Blues')
# 修改字体大小
plt.xticks(fontsize=12)
plt.yticks(fontsize=12)
plt.title('Confusion Matrix', fontsize=12)
plt.xlabel('Predicted Label', fontsize=12)
plt.ylabel('True Labels', fontsize=10)
# 调整布局防止重叠
plt.tight_layout()
# 显示图形
plt.show()
12、调用模型进行预测
# 调用模型进行预测
test_X = X_test[0].reshape(1, -1)
pred = model(test_X.to(device)).argmax(1).item()
print("模型预测结果为： ",pred)
print("="*20)
print("0: 未患病")
print("1: 已患病")
输出：
模型预测结果为： 0
====================
0: 未患病
1: 已患病