{"id":2421,"date":"2025-07-16T07:58:00","date_gmt":"2025-07-15T23:58:00","guid":{"rendered":"https:\/\/blog.laoyulaoyu.top\/?p=2421"},"modified":"2025-07-09T18:09:58","modified_gmt":"2025-07-09T10:09:58","slug":"%e6%88%91%e8%ae%a915%e4%b8%aaai%e6%a8%a1%e5%9e%8b%e6%89%93%e6%93%82%e5%8f%b0%ef%bc%8c%e9%a2%84%e6%b5%8b%e8%82%a1%e4%bb%b7%e8%af%af%e5%b7%ae%e4%bb%850-89%ef%bc%81%ef%bc%88%e9%99%84%e5%ae%9e%e6%88%98","status":"publish","type":"post","link":"https:\/\/laoyulaoyu.com\/index.php\/2025\/07\/16\/%e6%88%91%e8%ae%a915%e4%b8%aaai%e6%a8%a1%e5%9e%8b%e6%89%93%e6%93%82%e5%8f%b0%ef%bc%8c%e9%a2%84%e6%b5%8b%e8%82%a1%e4%bb%b7%e8%af%af%e5%b7%ae%e4%bb%850-89%ef%bc%81%ef%bc%88%e9%99%84%e5%ae%9e%e6%88%98\/","title":{"rendered":"\u6211\u8ba915\u4e2aAI\u6a21\u578b\u6253\u64c2\u53f0\uff0c\u9884\u6d4b\u80a1\u4ef7\u8bef\u5dee\u4ec50.89%\uff01\uff08\u9644\u5b9e\u6218\u4ee3\u7801\uff09"},"content":{"rendered":"\n

\u4f5c \u8005\uff1a\u8001\u4f59\u635e\u9c7c<\/a><\/p>\n\n\n\n

\u539f\u521b\u4e0d\u6613\uff0c\u8f6c\u8f7d\u8bf7\u6807\u660e\u51fa\u5904\u53ca\u539f\u4f5c\u8005\u3002<\/mark><\/strong><\/p>\n\n\n\n

\"\"<\/figure>\n\n\n\n
\n
\u5199\u5728\u524d\u9762\u7684\u8bdd\uff1a<\/strong>\u6700\u8fd1\u6211\u6d4b\u8bd5\u4e86\u8ba9\u591a\u4e2aAI\u6a21\u578b\u7ade\u4e89\u9884\u6d4b\u80a1\u4ef7\u7684\u64cd\u4f5c\uff01\u80dc\u51fa\u8005\u9884\u6d4b\u82f9\u679c\u80a1\u4ef7\u8bef\u5dee\u4ec50.89%\u3002<\/mark>\u8fd9\u7bc7\u624b\u628a\u624b\u6559\u4f60\u590d\u5236\u8fd9\u5957\u65b9\u6cd5\uff0c\u4ece\u6570\u636e\u83b7\u53d6\u5230\u6a21\u578bPK\u5168\u6d41\u7a0b\u8be6\u89e3\uff0c\u9644\u5b9e\u6218\u4ee3\u7801\u548c\u6548\u679c\u5bf9\u6bd4\u3002\u4ece\u6b64\u544a\u522b\u778e\u731c\uff0c\u7528\u79d1\u5b66\u65b9\u6cd5\u505a\u6295\u8d44\u51b3\u7b56\uff01<\/pre>\n<\/blockquote>\n\n\n\n
\u4e00\u3001\u53ea\u76ef\u4e00\u4e2a\u6a21\u578b\u4e0d\u9760\u8c31\uff1f<\/strong><\/h2>\n\n\n\n
\u80a1\u5e02\u662f\u4e2a\u5927\u821e\u53f0\uff0c\u6bcf\u5929\u90fd\u6709\u65e0\u6570\u56e0\u7d20\u5728\u5f71\u54cd\u80a1\u4ef7\u7684\u8d77\u4f0f\u3002\u65b0\u95fb\u3001\u653f\u7b56\u3001\u751a\u81f3\u6295\u8d44\u8005\u7684\u60c5\u7eea\uff0c\u90fd\u53ef\u80fd\u8ba9\u80a1\u4ef7\u5ffd\u4e0a\u5ffd\u4e0b\u3002\u4f5c\u4e3a\u6295\u8d44\u8005\uff0c\u8c01\u4e0d\u60f3\u63d0\u524d\u77e5\u9053\u80a1\u4ef7\u7684\u8d70\u52bf\uff0c\u505a\u51fa\u66f4\u806a\u660e\u7684\u6295\u8d44\u51b3\u5b9a\u5462\uff1f\u4f46\u8bf4\u5b9e\u8bdd\uff0c\u9884\u6d4b\u80a1\u4ef7\u53ef\u6ca1\u90a3\u4e48\u7b80\u5355\u3002\u4f20\u7edf\u7684\u9884\u6d4b\u65b9\u6cd5\u901a\u5e38\u53ea\u7528\u4e00\u4e2a\u6a21\u578b\uff0c\u6bd4\u5982\u7ebf\u6027\u56de\u5f52\u6216\u795e\u7ecf\u7f51\u7edc\uff0c\u4f46\u8fd9\u4e9b\u6a21\u578b\u5f88\u96be\u5b8c\u5168\u6293\u4f4f\u5e02\u573a\u7684\u590d\u6742\u53d8\u5316\u3002<\/p>\n\n\n\n
\u90a3\u4e48\uff0c\u6709\u6ca1\u6709\u4e00\u79cd\u66f4\u9760\u8c31\u7684\u65b9\u6cd5\u5462\uff1f\u7b54\u6848\u5c31\u662f\u591a\u6a21\u578b\u5bf9\u6bd4+\u5b9e\u6d4b\u9a8c\u8bc1<\/strong>\u65b9\u6cd5\uff0c\u4f18\u80dc\u52a3\u6c70\u6311\u6700\u9760\u8c31\u7684\u7b97\u6cd5\u3002<\/p>\n\n\n\n
\u8fd9\u4e2a\u65b9\u6cd5\u4e0d\u4f9d\u8d56\u5355\u4e00\u6a21\u578b\uff0c\u800c\u662f\u8ba9\u591a\u4e2a\u6a21\u578b\u201c\u6bd4\u6b66\u201d\uff0c\u9009\u51fa\u8868\u73b0\u6700\u597d\u7684\u90a3\u4e2a\u6765\u9884\u6d4b\u80a1\u4ef7\u3002\u8fd9\u79cd\u65b9\u5f0f\u5c31\u50cf\u7ed9\u4f60\u7684\u9884\u6d4b\u5de5\u5177\u7bb1\u88c5\u4e0a\u591a\u79cd\u6b66\u5668\uff0c\u54ea\u4e2a\u6700\u597d\u7528\u5c31\u7528\u54ea\u4e2a\uff01<\/p>\n\n\n\n
\u63a5\u4e0b\u6765\uff0c\u6211\u5c06\u5e26\u4f60\u7528Python\u5b9e\u73b0\u8fd9\u4e2a\u65b9\u6cd5\uff0c\u5e76\u4ee5\u82f9\u679c\u516c\u53f8\uff08AAPL\uff09\u7684\u80a1\u7968\u4e3a\u4f8b\uff0c\u5c55\u793a\u5b83\u7684\u5a01\u529b\u3002<\/p>\n\n\n\n
\u4e8c\u3001\u591a\u6a21\u578b\u65b9\u6cd5\u7684\u6838\u5fc3\u601d\u8def<\/strong><\/h2>\n\n\n\n
\u591a\u6a21\u578b\u65b9\u6cd5\u7684\u6838\u5fc3\u5728\u4e8e\u591a\u6837\u6027\u548c\u7cfb\u7edf\u6027<\/strong>\u3002\u6211\u4eec\u4f1a\u51c6\u5907\u4e00\u5927\u5806\u673a\u5668\u5b66\u4e60\u6a21\u578b\uff0c\u7136\u540e\u628a\u6570\u636e\u5206\u6210\u4e24\u90e8\u5206\uff1a\u8bad\u7ec3\u96c6\u7528\u6765\u6559\u6a21\u578b\uff0c\u6d4b\u8bd5\u96c6\u7528\u6765\u8003\u6a21\u578b\u3002\u6211\u4eec\u8fd8\u4f1a\u7528\u4e00\u4e9b\u6307\u6807\uff08\u6bd4\u5982\u5747\u65b9\u6839\u8bef\u5dee\uff0c\u7b80\u79f0RMSE\uff09\u6765\u6bd4\u8f83\u6bcf\u4e2a\u6a21\u578b\u7684\u8868\u73b0\uff0c\u6700\u540e\u9009\u51fa\u201c\u6210\u7ee9\u201d\u6700\u597d\u7684\u6a21\u578b\u3002\u6838\u5fc3\u601d\u8def\u903b\u8f91\u5982\u4e0b\uff1a<\/p>\n\n\n\n
    \n
  • \u7528\u9009\u4e2d\u7684\u6a21\u578b\u9884\u6d4b\u672a\u6765\u51e0\u5929<\/strong>\uff1a\u7ed3\u5408\u6700\u65b0\u6570\u636e\u6253\u5305\u9884\u6d4b\u3002<\/li>\n\n\n\n
  • \u51c6\u5907\u591a\u79cd\u6a21\u578b<\/strong>\uff1a\u7ebf\u6027\u56de\u5f52\u3001Lasso\/CatBoost\/XGBoost\u3001\u968f\u673a\u68ee\u6797\u7b49\uff1b<\/li>\n\n\n\n
  • \u6309\u5386\u53f2\u6570\u636e\u8bc4\u4f30<\/strong>\uff1a\u5bf9\u6bd4\u5404\u6a21\u578b\u7684MAE\u3001RMSE\u3001R\u00b2\u6307\u6807\uff1b<\/li>\n\n\n\n
  • \u9488\u5bf9\u7279\u5b9a\u5173\u952e\u65e5\u671f\u505a\u4e00\u6b21\u9884\u6d4b\u6d4b\u8bd5<\/strong>\uff1a\u770b\u6a21\u578b\u5728\u5173\u952e\u65f6\u70b9\u9884\u6d4b\u51c6\u4e0d\u51c6\uff1b<\/li>\n\n\n\n
  • \u9009\u51fa\u6700\u7a33\u6a21\u578b<\/strong>\uff1a\u901a\u8fc7\u504f\u5dee\u548c\u547d\u4e2d\u7684\u8868\u73b0\u51b3\u5b9a\u7528\u8c01\u3002<\/li>\n<\/ul>\n\n\n\n
    \u4f46\u5149\u5728\u6d4b\u8bd5\u96c6\u4e0a\u8868\u73b0\u597d\u8fd8\u4e0d\u591f\uff01\u6211\u4eec\u8fd8\u9700\u8981\u5728\u4e00\u4e9b\u5173\u952e\u65e5\u671f\u4e0a\u9a8c\u8bc1\u6a21\u578b\u7684\u9884\u6d4b\u80fd\u529b\uff0c\u786e\u4fdd\u5b83\u5728\u771f\u5b9e\u573a\u666f\u4e2d\u4e5f\u80fd\u9760\u8c31\u3002\u8fd9\u5c31\u50cf\u4e0d\u4ec5\u8981\u770b\u8003\u8bd5\u6210\u7ee9\uff0c\u8fd8\u8981\u770b\u5b83\u5728\u201c\u5b9e\u6218\u201d\u4e2d\u7684\u8868\u73b0\u3002<\/p>\n\n\n\n
    \u4e09\u3001\u624b\u628a\u624b\u4ee3\u7801\u6559\u5b66<\/strong><\/h2>\n\n\n\n
    \u4e0b\u9762\u8ba9\u6211\u4eec\u7b80\u7b80\u5355\u5355\u75284\u6b65\u6765\u641e\u5b9a\uff01\u8fd9\u4e2aPython\u7a0b\u5e8f\u4f1a\u5e2e\u6211\u4eec\u4ece\u83b7\u53d6\u6570\u636e\u5230\u9884\u6d4b\u80a1\u4ef7\uff0c\u5168\u7a0b\u81ea\u52a8\u5316\u5b8c\u6210\u3002<\/p>\n\n\n\n
    \u7b2c\u4e00\u6b65\uff1a\u83b7\u53d6\u5b9e\u65f6\u6570\u636e<\/strong><\/h3>\n\n\n\n
    \u8fd9\u4e00\u6b65\u662f\u4eceYahoo Finance\u4e0b\u8f7d\u6307\u5b9a\u80a1\u7968\u7684\u5386\u53f2\u6570\u636e\uff0c\u6bd4\u5982\u5f00\u76d8\u4ef7\u3001\u6536\u76d8\u4ef7\u3001\u6210\u4ea4\u91cf\u7b49\u3002\u6211\u4eec\u7528\u82f9\u679c\u516c\u53f8\uff08AAPL\uff09\u4f5c\u4e3a\u4f8b\u5b50\uff0c\u83b7\u53d62020\u5e74\u52302025\u5e74\u7684\u6570\u636e\u3002<\/p>\n\n\n\n
    def fetch_real_data(ticker, start_date, end_date):\n    # \u4f7f\u7528yfinance\u83b7\u53d6\u80a1\u7968\u6570\u636e\n    df = yf.download(ticker, start=start_date, end=end_date)\n    return df<\/code><\/pre>\n\n\n\n
    \u7b2c\u4e8c\u6b65\uff1a\u52a0\u5de5\u5173\u952e\u7279\u5f81<\/strong><\/h3>\n\n\n\n
    \u628a\u539f\u59cb\u6570\u636e\u53d8\u6210AI\u80fd\u770b\u61c2\u7684\u8bed\u8a00\uff08\u628a\u539f\u59cb\u6570\u636e\u52a0\u5de5\u6210\u6a21\u578b\u80fd\u7406\u89e3\u7684\u683c\u5f0f\uff09\u3002\u6211\u4eec\u6dfb\u52a0\u4e86\u65f6\u95f4\u7279\u5f81\uff08\u65e5\u3001\u6708\u3001\u5e74\uff09\uff0c\u5e76\u628a\u4e0b\u4e00\u5929\u7684\u6536\u76d8\u4ef7\u8bbe\u4e3a\u76ee\u6807\u53d8\u91cf\u3002<\/p>\n\n\n\n
    def prepare_features_basic(df):\n    # \u6dfb\u52a0\u65f6\u95f4\u7279\u5f81\n    df['Day'] = df.index.day\n    df['Month'] = df.index.month\n    df['Year'] = df.index.year\n    # \u521b\u5efa\u76ee\u6807\u53d8\u91cf\uff1a\u4e0b\u4e00\u5929\u7684\u6536\u76d8\u4ef7\n    df['Target'] = df['Close'].shift(-1)\n    # \u79fb\u9664\u6700\u540e\u4e00\u884c\uff0c\u56e0\u4e3a\u5b83\u6ca1\u6709\u76ee\u6807\u503c\n    df = df.iloc[:-1]\n    return df<\/code><\/pre>\n\n\n\n
    \u7b2c\u4e09\u6b65\uff1a\u6a21\u578b\u64c2\u53f0\u8d5b\uff08\u6838\u5fc3\uff01\uff09<\/strong><\/h3>\n\n\n\n
    \u8ba9\u591a\u4e2a\u4e3b\u6d41\u7b97\u6cd5\u540c\u573a\u7ade\u6280\uff0c\u8ba9\u5b83\u4eec\u5728\u8bad\u7ec3\u96c6\u4e0a\u5b66\u4e60\uff0c\u7136\u540e\u5728\u6d4b\u8bd5\u96c6\u4e0a\u9884\u6d4b\u80a1\u4ef7\u3002\u8fd8\u4f1a\u5bf9\u6bcf\u4e2a\u6a21\u578b\u8fdb\u884c\u8bc4\u5206\u3002<\/p>\n\n\n\n
    from sklearn.ensemble import RandomForestRegressor\nfrom xgboost import XGBRegressor\nimport lightgbm as lgb\n\n# \u53c2\u8d5b\u9009\u624b\u540d\u5355\nmodels = {\n    \"\u968f\u673a\u68ee\u6797\": RandomForestRegressor(),\n    \"XGBoost\": XGBRegressor(),\n    \"LightGBM\": lgb.LGBMRegressor(),\n    \"\u7ebf\u6027\u56de\u5f52\": LinearRegression(),\n    #...\u5176\u4ed6\u6a21\u578b\n}\n\n# \u8bc4\u5206\u89c4\u5219\uff1a\u8bef\u5dee\u8d8a\u5c0f\u5206\u8d8a\u9ad8\nfor name, model in models.items():\n    model.fit(X_train, y_train) # \u8bad\u7ec3\n    pred = model.predict(X_test) # \u9884\u6d4b\n    score = mean_squared_error(y_test, pred) # \u8bc4\u5206<\/code><\/pre>\n\n\n\n
    \u7b2c\u56db\u6b65\uff1a\u5b9e\u6218\u68c0\u9a8c<\/strong><\/h3>\n\n\n\n
    \u91cd\u70b9\u6765\u4e86\uff01\u6a21\u62df\u771f\u5b9e\u573a\u666f\uff1a\u53ea\u7528\u9884\u6d4b\u65e5\u4e4b\u524d\u7684\u6570\u636e\u6765\u8bad\u7ec3\u6a21\u578b\uff0c\u9884\u6d4b\u5f53\u5929\u7684\u6536\u76d8\u4ef7\uff0c\u7136\u540e\u548c\u5b9e\u9645\u4ef7\u683c\u5bf9\u6bd4\uff0c\u7b97\u51fa\u8bef\u5dee\u3002<\/p>\n\n\n\n
    # \u6d4b\u8bd52023\u5e74\u80a1\u707e\u65e5\u8868\u73b0\ncrisis_dates = ['2023-03-10','2023-06-15'] \nfor date in crisis_dates:\n    # \u7528\u5386\u53f2\u6570\u636e\u9884\u6d4b\n    pred = best_model.predict(pre_date_data)\n    # \u5bf9\u6bd4\u5b9e\u9645\u80a1\u4ef7\n    error = abs(pred - real_price)<\/code><\/pre>\n\n\n\n
    \u5728\u6211\u4eec\u7684\u4f8b\u5b50\u4e2d\uff0cLasso Regression\u88ab\u9009\u4e3a\u6700\u4f73\u6a21\u578b\u3002\u6211\u4eec\u57282025\u5e745\u670830\u65e5\u30016\u67082\u65e5\u30016\u67083\u65e5\u548c6\u67084\u65e5\u8fd9\u51e0\u4e2a\u65e5\u671f\u4e0a\u9a8c\u8bc1\u4e86\u5b83\u7684\u9884\u6d4b\u6548\u679c\u3002\u7ed3\u679c\u5982\u4e0b\uff1a<\/p>\n\n\n\n
    \u9884\u6d4b\u65e5\u671f<\/th>\u9884\u6d4b\u6536\u76d8\u4ef7<\/th>\u5b9e\u9645\u6536\u76d8\u4ef7<\/th>\u8bef\u5dee ($)<\/th>\u8bef\u5dee (%)<\/th><\/tr>
    2025-05-30<\/td>200.32<\/td>200.85<\/td>0.53<\/td>0.26<\/td><\/tr>
    2025-06-02<\/td>199.68<\/td>201.70<\/td>2.02<\/td>1.00<\/td><\/tr>
    2025-06-03<\/td>200.06<\/td>203.27<\/td>3.21<\/td>1.58<\/td><\/tr>
    2025-06-04<\/td>201.40<\/td>202.82<\/td>1.42<\/td>0.70<\/td><\/tr><\/tbody><\/table><\/figure>\n\n\n\n
    \u5206\u6790<\/strong>\uff1a\u5e73\u5747\u7edd\u5bf9\u8bef\u5dee\u4e3a1.79\u7f8e\u5143\uff0c\u5e73\u5747\u767e\u5206\u6bd4\u8bef\u5dee\u4e3a0.89%\uff0c75%\u7684\u9884\u6d4b\u8bef\u5dee\u57281%\u4ee5\u5185\u3002\u8fd9\u8bf4\u660e\u6a21\u578b\u5728\u8fd9\u4e9b\u65e5\u671f\u4e0a\u7684\u8868\u73b0\u76f8\u5f53\u4e0d\u9519\uff01<\/p>\n\n\n\n
    \u6211\u4eec\u8fd8\u7528\u6700\u4f73\u6a21\u578b\u9884\u6d4b\u4e86\u672a\u67655\u4e2a\u4ea4\u6613\u65e5\u7684\u6536\u76d8\u4ef7\uff1a<\/p>\n\n\n\n
    \u65e5\u671f<\/th>\u9884\u6d4b\u6536\u76d8\u4ef7<\/th><\/tr>
    2025-06-05<\/td>202.66<\/td><\/tr>
    2025-06-06<\/td>202.65<\/td><\/tr>
    2025-06-09<\/td>202.63<\/td><\/tr>
    2025-06-10<\/td>202.61<\/td><\/tr>
    2025-06-11<\/td>202.59<\/td><\/tr><\/tbody><\/table><\/figure>\n\n\n\n
    \u6280\u672f\u4eae\u70b9<\/strong><\/h3>\n\n\n\n