Compare commits
No commits in common. 'main' and 'master' have entirely different histories.
@ -0,0 +1,8 @@
|
|||||||
|
# Default ignored files
|
||||||
|
/shelf/
|
||||||
|
/workspace.xml
|
||||||
|
# Editor-based HTTP Client requests
|
||||||
|
/httpRequests/
|
||||||
|
# Datasource local storage ignored files
|
||||||
|
/dataSources/
|
||||||
|
/dataSources.local.xml
|
||||||
@ -0,0 +1,6 @@
|
|||||||
|
<component name="InspectionProjectProfileManager">
|
||||||
|
<settings>
|
||||||
|
<option name="USE_PROJECT_PROFILE" value="false" />
|
||||||
|
<version value="1.0" />
|
||||||
|
</settings>
|
||||||
|
</component>
|
||||||
@ -0,0 +1,7 @@
|
|||||||
|
<?xml version="1.0" encoding="UTF-8"?>
|
||||||
|
<project version="4">
|
||||||
|
<component name="Black">
|
||||||
|
<option name="sdkName" value="Python 3.13 (大作业)" />
|
||||||
|
</component>
|
||||||
|
<component name="ProjectRootManager" version="2" project-jdk-name="Python 3.13 (大作业)" project-jdk-type="Python SDK" />
|
||||||
|
</project>
|
||||||
@ -0,0 +1,8 @@
|
|||||||
|
<?xml version="1.0" encoding="UTF-8"?>
|
||||||
|
<project version="4">
|
||||||
|
<component name="ProjectModuleManager">
|
||||||
|
<modules>
|
||||||
|
<module fileurl="file://$PROJECT_DIR$/.idea/大作业.iml" filepath="$PROJECT_DIR$/.idea/大作业.iml" />
|
||||||
|
</modules>
|
||||||
|
</component>
|
||||||
|
</project>
|
||||||
@ -0,0 +1,6 @@
|
|||||||
|
<?xml version="1.0" encoding="UTF-8"?>
|
||||||
|
<project version="4">
|
||||||
|
<component name="VcsDirectoryMappings">
|
||||||
|
<mapping directory="$PROJECT_DIR$" vcs="Git" />
|
||||||
|
</component>
|
||||||
|
</project>
|
||||||
@ -0,0 +1,10 @@
|
|||||||
|
<?xml version="1.0" encoding="UTF-8"?>
|
||||||
|
<module type="PYTHON_MODULE" version="4">
|
||||||
|
<component name="NewModuleRootManager">
|
||||||
|
<content url="file://$MODULE_DIR$">
|
||||||
|
<excludeFolder url="file://$MODULE_DIR$/.venv" />
|
||||||
|
</content>
|
||||||
|
<orderEntry type="jdk" jdkName="Python 3.13 (大作业)" jdkType="Python SDK" />
|
||||||
|
<orderEntry type="sourceFolder" forTests="false" />
|
||||||
|
</component>
|
||||||
|
</module>
|
||||||
Binary file not shown.
Binary file not shown.
File diff suppressed because it is too large
Load Diff
Binary file not shown.
Binary file not shown.
Binary file not shown.
Binary file not shown.
Binary file not shown.
Binary file not shown.
Binary file not shown.
Binary file not shown.
@ -0,0 +1,192 @@
|
|||||||
|
import torch
|
||||||
|
import torch.nn as nn
|
||||||
|
import torch.optim as optim
|
||||||
|
import numpy as np
|
||||||
|
import pandas as pd
|
||||||
|
from sklearn.preprocessing import StandardScaler
|
||||||
|
import joblib
|
||||||
|
import os
|
||||||
|
import sys
|
||||||
|
|
||||||
|
# 添加项目根目录到Python路径
|
||||||
|
sys.path.append(os.path.dirname(os.path.dirname(os.path.abspath(__file__))))
|
||||||
|
|
||||||
|
class AutoEncoder(nn.Module):
|
||||||
|
"""
|
||||||
|
自编码器用于异常检测
|
||||||
|
"""
|
||||||
|
def __init__(self, input_dim, hidden_dim1, hidden_dim2):
|
||||||
|
super(AutoEncoder, self).__init__()
|
||||||
|
# 编码器
|
||||||
|
self.encoder = nn.Sequential(
|
||||||
|
nn.Linear(input_dim, hidden_dim1),
|
||||||
|
nn.ReLU(True),
|
||||||
|
nn.Linear(hidden_dim1, hidden_dim2),
|
||||||
|
nn.ReLU(True)
|
||||||
|
)
|
||||||
|
|
||||||
|
# 解码器
|
||||||
|
self.decoder = nn.Sequential(
|
||||||
|
nn.Linear(hidden_dim2, hidden_dim1),
|
||||||
|
nn.ReLU(True),
|
||||||
|
nn.Linear(hidden_dim1, input_dim),
|
||||||
|
nn.ReLU(True)
|
||||||
|
)
|
||||||
|
|
||||||
|
def forward(self, x):
|
||||||
|
x = self.encoder(x)
|
||||||
|
x = self.decoder(x)
|
||||||
|
return x
|
||||||
|
|
||||||
|
class AdversarialAutoEncoder(nn.Module):
|
||||||
|
"""
|
||||||
|
对抗自编码器
|
||||||
|
"""
|
||||||
|
def __init__(self, input_dim, hidden_dim1, hidden_dim2, latent_dim):
|
||||||
|
super(AdversarialAutoEncoder, self).__init__()
|
||||||
|
# 编码器
|
||||||
|
self.encoder = nn.Sequential(
|
||||||
|
nn.Linear(input_dim, hidden_dim1),
|
||||||
|
nn.ReLU(True),
|
||||||
|
nn.Linear(hidden_dim1, hidden_dim2),
|
||||||
|
nn.ReLU(True),
|
||||||
|
nn.Linear(hidden_dim2, latent_dim),
|
||||||
|
nn.ReLU(True)
|
||||||
|
)
|
||||||
|
|
||||||
|
# 解码器
|
||||||
|
self.decoder = nn.Sequential(
|
||||||
|
nn.Linear(latent_dim, hidden_dim2),
|
||||||
|
nn.ReLU(True),
|
||||||
|
nn.Linear(hidden_dim2, hidden_dim1),
|
||||||
|
nn.ReLU(True),
|
||||||
|
nn.Linear(hidden_dim1, input_dim),
|
||||||
|
nn.Sigmoid() # 使用Sigmoid确保输出在0-1之间
|
||||||
|
)
|
||||||
|
|
||||||
|
# 判别器
|
||||||
|
self.discriminator = nn.Sequential(
|
||||||
|
nn.Linear(latent_dim, hidden_dim2),
|
||||||
|
nn.ReLU(True),
|
||||||
|
nn.Linear(hidden_dim2, hidden_dim1),
|
||||||
|
nn.ReLU(True),
|
||||||
|
nn.Linear(hidden_dim1, 1),
|
||||||
|
nn.Sigmoid()
|
||||||
|
)
|
||||||
|
|
||||||
|
def encode(self, x):
|
||||||
|
return self.encoder(x)
|
||||||
|
|
||||||
|
def decode(self, z):
|
||||||
|
return self.decoder(z)
|
||||||
|
|
||||||
|
def discriminate(self, z):
|
||||||
|
return self.discriminator(z)
|
||||||
|
|
||||||
|
def forward(self, x):
|
||||||
|
z = self.encode(x)
|
||||||
|
recon_x = self.decode(z)
|
||||||
|
return recon_x, z
|
||||||
|
|
||||||
|
def train_adversarial_autoencoder():
|
||||||
|
"""
|
||||||
|
训练对抗自编码器
|
||||||
|
"""
|
||||||
|
# 读取数据
|
||||||
|
print("读取信贷数据...")
|
||||||
|
df = pd.read_csv('data/credit_data.csv')
|
||||||
|
|
||||||
|
# 只使用数值特征进行自编码器训练
|
||||||
|
numerical_features = ['age', 'income', 'employment_length', 'loan_amount',
|
||||||
|
'credit_score', 'debt_to_income', 'num_credit_lines']
|
||||||
|
X = df[numerical_features]
|
||||||
|
|
||||||
|
# 标准化数据
|
||||||
|
scaler = StandardScaler()
|
||||||
|
X_scaled = scaler.fit_transform(X)
|
||||||
|
|
||||||
|
# 转换为PyTorch张量
|
||||||
|
X_tensor = torch.FloatTensor(X_scaled)
|
||||||
|
|
||||||
|
# 设置模型参数
|
||||||
|
input_dim = X_tensor.shape[1]
|
||||||
|
hidden_dim1 = 64
|
||||||
|
hidden_dim2 = 32
|
||||||
|
latent_dim = 16
|
||||||
|
|
||||||
|
# 创建模型
|
||||||
|
model = AdversarialAutoEncoder(input_dim, hidden_dim1, hidden_dim2, latent_dim)
|
||||||
|
|
||||||
|
# 设置损失函数和优化器
|
||||||
|
reconstruction_criterion = nn.MSELoss()
|
||||||
|
adversarial_criterion = nn.BCELoss()
|
||||||
|
|
||||||
|
autoencoder_optimizer = optim.Adam(
|
||||||
|
list(model.encoder.parameters()) + list(model.decoder.parameters()),
|
||||||
|
lr=0.001
|
||||||
|
)
|
||||||
|
discriminator_optimizer = optim.Adam(model.discriminator.parameters(), lr=0.001)
|
||||||
|
|
||||||
|
# 训练模型
|
||||||
|
num_epochs = 100
|
||||||
|
batch_size = 64
|
||||||
|
|
||||||
|
print("开始训练对抗自编码器...")
|
||||||
|
for epoch in range(num_epochs):
|
||||||
|
for i in range(0, len(X_tensor), batch_size):
|
||||||
|
batch = X_tensor[i:i+batch_size]
|
||||||
|
|
||||||
|
# 训练自编码器
|
||||||
|
autoencoder_optimizer.zero_grad()
|
||||||
|
|
||||||
|
recon_batch, latent_batch = model(batch)
|
||||||
|
real_labels = torch.ones(batch.size(0), 1)
|
||||||
|
fake_labels = torch.zeros(batch.size(0), 1)
|
||||||
|
|
||||||
|
# 重构损失
|
||||||
|
recon_loss = reconstruction_criterion(recon_batch, batch)
|
||||||
|
|
||||||
|
# 对抗损失 - 生成器希望判别器将生成的潜在向量识别为真实
|
||||||
|
disc_fake = model.discriminate(latent_batch)
|
||||||
|
adversarial_loss = adversarial_criterion(disc_fake, real_labels)
|
||||||
|
|
||||||
|
autoencoder_loss = recon_loss + 0.1 * adversarial_loss
|
||||||
|
autoencoder_loss.backward()
|
||||||
|
autoencoder_optimizer.step()
|
||||||
|
|
||||||
|
# 训练判别器
|
||||||
|
discriminator_optimizer.zero_grad()
|
||||||
|
|
||||||
|
# 真实潜在向量（从标准正态分布采样）
|
||||||
|
real_latent = torch.randn(batch.size(0), latent_dim)
|
||||||
|
disc_real = model.discriminate(real_latent)
|
||||||
|
disc_real_loss = adversarial_criterion(disc_real, real_labels)
|
||||||
|
|
||||||
|
# 生成的潜在向量
|
||||||
|
disc_fake = model.discriminate(latent_batch.detach())
|
||||||
|
disc_fake_loss = adversarial_criterion(disc_fake, fake_labels)
|
||||||
|
|
||||||
|
discriminator_loss = disc_real_loss + disc_fake_loss
|
||||||
|
discriminator_loss.backward()
|
||||||
|
discriminator_optimizer.step()
|
||||||
|
|
||||||
|
if (epoch + 1) % 10 == 0:
|
||||||
|
print(f'Epoch [{epoch+1}/{num_epochs}], '
|
||||||
|
f'Recon Loss: {recon_loss.item():.4f}, '
|
||||||
|
f'Adversarial Loss: {adversarial_loss.item():.4f}, '
|
||||||
|
f'Discriminator Loss: {discriminator_loss.item():.4f}')
|
||||||
|
|
||||||
|
# 保存模型和标准化器
|
||||||
|
print("保存对抗自编码器模型...")
|
||||||
|
torch.save(model.state_dict(), 'models/adversarial_autoencoder.pth')
|
||||||
|
joblib.dump(scaler, 'models/ae_scaler.pkl')
|
||||||
|
|
||||||
|
return model, scaler
|
||||||
|
|
||||||
|
if __name__ == "__main__":
|
||||||
|
# 检查是否有可用的GPU
|
||||||
|
device = torch.device('cuda' if torch.cuda.is_available() else 'cpu')
|
||||||
|
print(f"使用设备: {device}")
|
||||||
|
|
||||||
|
model, scaler = train_adversarial_autoencoder()
|
||||||
|
print("对抗自编码器训练完成!")
|
||||||
@ -0,0 +1,93 @@
|
|||||||
|
import sys
|
||||||
|
import os
|
||||||
|
|
||||||
|
# 添加项目根目录到Python路径
|
||||||
|
sys.path.append(os.path.dirname(os.path.dirname(os.path.abspath(__file__))))
|
||||||
|
|
||||||
|
import pandas as pd
|
||||||
|
import numpy as np
|
||||||
|
from sklearn.model_selection import train_test_split
|
||||||
|
from sklearn.metrics import accuracy_score, classification_report, confusion_matrix
|
||||||
|
import lightgbm as lgb
|
||||||
|
import joblib
|
||||||
|
|
||||||
|
from data.data_generator import preprocess_data
|
||||||
|
|
||||||
|
def train_lightgbm_model():
|
||||||
|
"""
|
||||||
|
训练LightGBM模型用于信贷风险评估
|
||||||
|
"""
|
||||||
|
# 读取数据
|
||||||
|
print("读取信贷数据...")
|
||||||
|
df = pd.read_csv('data/credit_data.csv')
|
||||||
|
print(f"数据形状: {df.shape}")
|
||||||
|
|
||||||
|
# 数据预处理
|
||||||
|
print("数据预处理...")
|
||||||
|
X, y, scaler, le_education, le_home, le_purpose = preprocess_data(df)
|
||||||
|
|
||||||
|
# 分割数据集
|
||||||
|
X_train, X_test, y_train, y_test = train_test_split(
|
||||||
|
X, y, test_size=0.2, random_state=42, stratify=y
|
||||||
|
)
|
||||||
|
|
||||||
|
print(f"训练集大小: {X_train.shape}")
|
||||||
|
print(f"测试集大小: {X_test.shape}")
|
||||||
|
|
||||||
|
# 创建LightGBM分类器
|
||||||
|
print("创建LightGBM模型...")
|
||||||
|
model = lgb.LGBMClassifier(
|
||||||
|
n_estimators=200,
|
||||||
|
max_depth=8,
|
||||||
|
learning_rate=0.05,
|
||||||
|
num_leaves=64,
|
||||||
|
subsample=0.8,
|
||||||
|
colsample_bytree=0.8,
|
||||||
|
random_state=42,
|
||||||
|
verbose=-1
|
||||||
|
)
|
||||||
|
|
||||||
|
# 训练模型
|
||||||
|
print("训练模型...")
|
||||||
|
model.fit(X_train, y_train,
|
||||||
|
eval_set=[(X_test, y_test)],
|
||||||
|
eval_metric='binary_logloss',
|
||||||
|
callbacks=[lgb.early_stopping(10), lgb.log_evaluation(10)])
|
||||||
|
|
||||||
|
# 预测
|
||||||
|
print("模型预测...")
|
||||||
|
y_pred = model.predict(X_test)
|
||||||
|
y_pred_proba = model.predict_proba(X_test)[:, 1]
|
||||||
|
|
||||||
|
# 评估模型
|
||||||
|
accuracy = accuracy_score(y_test, y_pred)
|
||||||
|
print(f"模型准确率: {accuracy:.4f}")
|
||||||
|
|
||||||
|
print("\n分类报告:")
|
||||||
|
print(classification_report(y_test, y_pred))
|
||||||
|
|
||||||
|
print("\n混淆矩阵:")
|
||||||
|
print(confusion_matrix(y_test, y_pred))
|
||||||
|
|
||||||
|
# 保存模型和预处理器
|
||||||
|
print("保存模型和预处理器...")
|
||||||
|
joblib.dump(model, 'models/lightgbm_model.pkl')
|
||||||
|
joblib.dump(scaler, 'models/scaler.pkl')
|
||||||
|
joblib.dump(le_education, 'models/le_education.pkl')
|
||||||
|
joblib.dump(le_home, 'models/le_home.pkl')
|
||||||
|
joblib.dump(le_purpose, 'models/le_purpose.pkl')
|
||||||
|
|
||||||
|
# 特征重要性
|
||||||
|
feature_importance = pd.DataFrame({
|
||||||
|
'feature': X.columns,
|
||||||
|
'importance': model.feature_importances_
|
||||||
|
}).sort_values('importance', ascending=False)
|
||||||
|
|
||||||
|
print("\n特征重要性:")
|
||||||
|
print(feature_importance)
|
||||||
|
|
||||||
|
return model, scaler, le_education, le_home, le_purpose
|
||||||
|
|
||||||
|
if __name__ == "__main__":
|
||||||
|
model, scaler, le_education, le_home, le_purpose = train_lightgbm_model()
|
||||||
|
print("\n模型训练完成!")
|
||||||
@ -0,0 +1,88 @@
|
|||||||
|
import sys
|
||||||
|
import os
|
||||||
|
|
||||||
|
# 添加项目根目录到Python路径
|
||||||
|
sys.path.append(os.path.dirname(os.path.dirname(os.path.abspath(__file__))))
|
||||||
|
|
||||||
|
import pandas as pd
|
||||||
|
import numpy as np
|
||||||
|
from sklearn.model_selection import train_test_split
|
||||||
|
from sklearn.metrics import accuracy_score, classification_report, confusion_matrix
|
||||||
|
import xgboost as xgb
|
||||||
|
import joblib
|
||||||
|
|
||||||
|
from data.data_generator import preprocess_data
|
||||||
|
|
||||||
|
def train_xgboost_model():
|
||||||
|
"""
|
||||||
|
训练XGBoost模型用于信贷风险评估
|
||||||
|
"""
|
||||||
|
# 读取数据
|
||||||
|
print("读取信贷数据...")
|
||||||
|
df = pd.read_csv('data/credit_data.csv')
|
||||||
|
print(f"数据形状: {df.shape}")
|
||||||
|
|
||||||
|
# 数据预处理
|
||||||
|
print("数据预处理...")
|
||||||
|
X, y, scaler, le_education, le_home, le_purpose = preprocess_data(df)
|
||||||
|
|
||||||
|
# 分割数据集
|
||||||
|
X_train, X_test, y_train, y_test = train_test_split(
|
||||||
|
X, y, test_size=0.2, random_state=42, stratify=y
|
||||||
|
)
|
||||||
|
|
||||||
|
print(f"训练集大小: {X_train.shape}")
|
||||||
|
print(f"测试集大小: {X_test.shape}")
|
||||||
|
|
||||||
|
# 创建XGBoost分类器
|
||||||
|
print("创建XGBoost模型...")
|
||||||
|
model = xgb.XGBClassifier(
|
||||||
|
n_estimators=100,
|
||||||
|
max_depth=6,
|
||||||
|
learning_rate=0.1,
|
||||||
|
subsample=0.8,
|
||||||
|
colsample_bytree=0.8,
|
||||||
|
random_state=42
|
||||||
|
)
|
||||||
|
|
||||||
|
# 训练模型
|
||||||
|
print("训练模型...")
|
||||||
|
model.fit(X_train, y_train)
|
||||||
|
|
||||||
|
# 预测
|
||||||
|
print("模型预测...")
|
||||||
|
y_pred = model.predict(X_test)
|
||||||
|
y_pred_proba = model.predict_proba(X_test)[:, 1]
|
||||||
|
|
||||||
|
# 评估模型
|
||||||
|
accuracy = accuracy_score(y_test, y_pred)
|
||||||
|
print(f"模型准确率: {accuracy:.4f}")
|
||||||
|
|
||||||
|
print("\n分类报告:")
|
||||||
|
print(classification_report(y_test, y_pred))
|
||||||
|
|
||||||
|
print("\n混淆矩阵:")
|
||||||
|
print(confusion_matrix(y_test, y_pred))
|
||||||
|
|
||||||
|
# 保存模型和预处理器
|
||||||
|
print("保存模型和预处理器...")
|
||||||
|
joblib.dump(model, 'models/xgboost_model.pkl')
|
||||||
|
joblib.dump(scaler, 'models/scaler.pkl')
|
||||||
|
joblib.dump(le_education, 'models/le_education.pkl')
|
||||||
|
joblib.dump(le_home, 'models/le_home.pkl')
|
||||||
|
joblib.dump(le_purpose, 'models/le_purpose.pkl')
|
||||||
|
|
||||||
|
# 特征重要性
|
||||||
|
feature_importance = pd.DataFrame({
|
||||||
|
'feature': X.columns,
|
||||||
|
'importance': model.feature_importances_
|
||||||
|
}).sort_values('importance', ascending=False)
|
||||||
|
|
||||||
|
print("\n特征重要性:")
|
||||||
|
print(feature_importance)
|
||||||
|
|
||||||
|
return model, scaler, le_education, le_home, le_purpose
|
||||||
|
|
||||||
|
if __name__ == "__main__":
|
||||||
|
model, scaler, le_education, le_home, le_purpose = train_xgboost_model()
|
||||||
|
print("\n模型训练完成!")
|
||||||
Binary file not shown.
|
After Width: | Height: | Size: 56 KiB |
|
After Width: | Height: | Size: 269 KiB |
|
After Width: | Height: | Size: 57 KiB |
|
After Width: | Height: | Size: 58 KiB |
|
After Width: | Height: | Size: 52 KiB |
|
After Width: | Height: | Size: 57 KiB |
|
After Width: | Height: | Size: 88 KiB |
|
After Width: | Height: | Size: 58 KiB |
|
|
After Width: | Height: | Size: 87 KiB |
|
After Width: | Height: | Size: 53 KiB |
|
After Width: | Height: | Size: 54 KiB |
|
After Width: | Height: | Size: 112 KiB |
|
After Width: | Height: | Size: 263 KiB |
Loading…
Reference in new issue