第19讲：过拟合的危险 (The Dangers of Overfitting) · 练习

import yfinance as yf
import numpy as np
from sklearn.preprocessing import PolynomialFeatures
from sklearn.metrics import mean_absolute_percentage_error

# 数据获取与处理
msft = yf.download(...)[...]
dates = (...)
prices = (...)

# 划分训练集和测试集
split_idx = int(len(dates)*0.8)
X_train, X_test = (...)
y_train, y_test = (...)

# 多项式特征转换
poly3 = PolynomialFeatures(degree=3)
X_poly3 = poly3.fit_transform(...)
poly9 = PolynomialFeatures(degree=9)
X_poly9 = poly9.fit_transform(...)

# 模型训练
model3 = LinearRegression().fit(...)
model9 = (...)

from sklearn.linear_model import Ridge, Lasso

# 获取多资产数据
symbols = ['AAPL','MSFT','GOOG','SPY',...]
data = yf.download(...)[...]
returns = (...)

# 设置特征和标签
X = (...)
y = (...)

# 划分数据集
X_train, X_test, y_train, y_test = (...)

# 模型训练
models = {
    'OLS': LinearRegression(),
    'Ridge': Ridge(alpha=0.5),
    'Lasso': Lasso(alpha=0.1)
}

for name, model in models.items():
    ...

# 系数对比可视化
...