{"id":3496,"date":"2026-01-29T07:59:00","date_gmt":"2026-01-28T23:59:00","guid":{"rendered":"https:\/\/www.laoyulaoyu.com\/?p=3496"},"modified":"2026-01-21T21:01:59","modified_gmt":"2026-01-21T13:01:59","slug":"%e4%bb%8e%e6%95%b0%e6%8d%ae%e4%b8%ad%e5%af%bb%e6%89%be%e7%a1%ae%e5%ae%9a%e6%80%a7%ef%bc%9a%e9%87%8f%e5%8c%96%e9%ab%98%e6%89%8b%e5%a6%82%e4%bd%95%e7%94%a8%e4%b8%89%e6%8b%9b%e5%88%92%e5%88%86%e8%a1%8c","status":"publish","type":"post","link":"https:\/\/laoyulaoyu.com\/index.php\/2026\/01\/29\/%e4%bb%8e%e6%95%b0%e6%8d%ae%e4%b8%ad%e5%af%bb%e6%89%be%e7%a1%ae%e5%ae%9a%e6%80%a7%ef%bc%9a%e9%87%8f%e5%8c%96%e9%ab%98%e6%89%8b%e5%a6%82%e4%bd%95%e7%94%a8%e4%b8%89%e6%8b%9b%e5%88%92%e5%88%86%e8%a1%8c\/","title":{"rendered":"\u4ece\u6570\u636e\u4e2d\u5bfb\u627e\u786e\u5b9a\u6027\uff1a\u91cf\u5316\u9ad8\u624b\u5982\u4f55\u7528\u4e09\u62db\u5212\u5206\u884c\u60c5\u8fd0\u884c\u533a\u95f4\uff1f"},"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>\u5f88\u591a\u4eba\u6700\u5bb9\u6613\u72af\u7684\u9519\u8bef\uff0c\u5c31\u662f\u62ff\u4e00\u5957\u56fa\u5b9a\u7684\u65b9\u6cd5\u53bb\u786c\u5957\u5343\u53d8\u4e07\u5316\u7684\u5e02\u573a\u3002\u8fd9\u7bc7\u6587\u7ae0\u6211\u5c06\u5206\u4eab\u4e09\u79cd\u8bc6\u522b\u884c\u60c5\u72b6\u6001\u7684\u786c\u6838\u65b9\u6cd5<\/mark>\uff1a\u4ece\u57fa\u7840\u7684\u5747\u7ebf\u903b\u8f91\uff0c\u5230\u6ce2\u52a8\u7684\u7edf\u8ba1\u89c4\u5f8b\uff0c\u518d\u5230AI\u5e38\u7528\u7684\u6570\u5b66\u6a21\u578b\u3002\u6211\u4f1a\u914d\u4e0a\u5177\u4f53\u7684\u903b\u8f91\u4ee3\u7801\u548c\u56fe\u8868\uff0c\u5e26\u4f60\u50cf\u6c14\u8c61\u5458\u4e00\u6837\u9884\u5224\u5e02\u573a\u73af\u5883\uff0c\u505a\u5230\u987a\u52bf\u800c\u4e3a\u3002<\/pre>\n<\/blockquote>\n\n\n\n
\u603b\u6709\u4eba\u5728\u731c\uff1a”\u8fd9\u5e02\u573a\u5230\u5e95\u662f\u5728\u84c4\u52bf\u5f85\u53d1\uff0c\u8fd8\u662f\u5728\u5077\u5077\u57cb\u96f7\uff1f”\u8bf4\u5b9e\u8bdd\uff0c\u6211\u8981\u662f\u80fd\u4e00\u773c\u770b\u7a7f\uff0c\u65e9\u5c31****\uff08\u54b3\u54b3\uff0c\u5408\u89c4\u5408\u89c4<\/em>\uff09\u3002\u4e0d\u8fc7\u8bdd\u8bf4\u56de\u6765\uff0c\u5e72\u4e86\u8fd9\u4e48\u591a\u5e74\u91cf\u5316\uff0c\u8fd8\u771f\u6709\u51e0\u62db\u80fd\u5e2e\u5927\u5bb6”\u542c\u542c\u5e02\u573a\u7684\u547c\u5438\u58f0”\u3002<\/p>\n\n\n\n
\u4eca\u5929\u4e0d\u804a\u865a\u7684\uff0c\u76f4\u63a5\u4e0a\u786c\u6838\u5e72\u8d27\u2014\u2014\u5e02\u573a\u72b6\u6001\u8bc6\u522b<\/strong>\u3002\u8fd9\u4e0d\u662f\u4ec0\u4e48\u7b97\u547d\u795e\u5668\uff0c\u800c\u662f\u6b63\u7ecf\u7684\u7edf\u8ba1\u5b66\u65b9\u6cd5\uff0c\u534e\u5c14\u8857\u90a3\u5e2e\u91cf\u5316\u56e2\u961f\u5929\u5929\u7422\u78e8\u7684\u4e1c\u897f\u3002\u6211\u628a\u5b83\u63b0\u5f00\u63c9\u788e\uff0c\u7528\u4e09\u62db\u6559\u4f60\u8bc6\u522b\u5e02\u573a\u5230\u5e95\u5728”\u88c5\u7761”\u8fd8\u662f”\u771f\u9192”\u3002<\/p>\n\n\n\n
\u2460 \u4e3a\u4ec0\u4e48\u8981\u8bc6\u522b<\/strong>\u5e02\u573a<\/strong>\u72b6\u6001\uff1f<\/strong><\/h2>\n\n\n\n
\u5e02\u573a\u72b6\u6001\u662f\u4e2a\u5565\u73a9\u610f\u513f\uff1f\u7b80\u5355\u8bf4\uff0c\u5e02\u573a\u72b6\u6001<\/strong>\u5c31\u662f\u5e02\u573a\u5728\u4e00\u6bb5\u65f6\u95f4\u5185\u7684”\u6027\u683c\u7279\u5f81<\/strong>“\u3002\u5c31\u50cf\u4eba\u6709\u559c\u6012\u54c0\u4e50\uff0c\u5e02\u573a\u4e5f\u6709\u4e09\u79cd\u57fa\u672c\u9762\u5b54\uff1a<\/p>\n\n\n\n
1.\u00a0\u8d8b\u52bf\u578b\u6027\u683c<\/strong>\u2014\u2014\u8981\u4e48\u4e00\u8def\u5411\u5317\uff0c\u8981\u4e48\u4e00\u8def\u5411\u5357\uff0c\u76ee\u6807\u660e\u786e\u3002<\/p>\n\n\n\n
2.\u00a0\u9707\u8361\u578b\u6027\u683c<\/strong>\u2014\u2014\u50cf\u4e2a\u949f\u6446\uff0c\u6765\u56de\u6643\u60a0\uff0c\u6709\u6839\u770b\u4e0d\u89c1\u7684\u7ef3\u5b50\u7275\u7740\u3002<\/p>\n\n\n\n
3.\u00a0\u66b4\u8e81\u578b\u6027\u683c<\/strong>\u2014\u2014\u4e0a\u8e7f\u4e0b\u8df3\uff0c\u4e00\u70b9\u5c31\u7740\uff0c\u98ce\u9669\u4f20\u67d3\u7279\u522b\u5feb\u3002<\/p>\n\n\n\n
\"\"\/<\/figure>\n\n\n\n
\u56fe1\uff1a\u6807\u666e500\u4e09\u79cd\u5e02\u573a\u72b6\u6001\u793a\u4f8b<\/em><\/p>\n\n\n\n
\u8bc6\u522b\u8fd9\u4e9b\u72b6\u6001\u6709\u5565\u7528\uff1f\u6838\u5fc3\u4f5c\u7528\u5c31\u4e24\u4e2a\uff1a\u89c4\u907f\u66b4\u8e81\u671f\u7684\u98ce\u9669\uff0c\u6293\u4f4f\u8d8b\u52bf\u671f\u7684\u7a97\u53e3<\/strong>\u3002\u5e02\u573a\u5d29\u76d8\u524d\u80fd\u7528\u8fd9\u62db\u907f\u5f00\u5927\u90e8\u5206\u51b2\u51fb\u7684\u670b\u53cb\uff0c\u665a\u4e0a\u7761\u5f97\u90fd\u9999\u3002<\/p>\n\n\n\n
\u2461 \u6570\u636e\u51c6\u5907\uff1a\u78e8\u5200\u4e0d\u8bef\u780d\u67f4\u5de5<\/strong><\/h2>\n\n\n\n

\u4e3a\u5565\u8981\u7528\u5bf9\u6570\u6536\u76ca\u7387\uff1f<\/strong><\/h3>\n\n\n\n
\u5f88\u591a\u65b0\u624b\u76f4\u63a5\u7528\u4ef7\u683c\u6da8\u8dcc\u7b97\u6536\u76ca\uff0c\u8fd9\u65b9\u6cd5\u4e0d\u662f\u4e0d\u884c\uff0c\u5c31\u662f\u6709\u70b9”\u7c97\u7cd9”\u3002\u5bf9\u6570\u6536\u76ca\u7387\u6709\u51e0\u4e2a\u597d\u5904\uff1a<\/p>\n\n\n\n
    \n
  • \u5bf9\u79f0\u6027\uff1a\u6da810%\u518d\u8dcc10%\uff0c\u4e0d\u4f1a\u56de\u5230\u539f\u70b9\uff0c\u5bf9\u6570\u6536\u76ca\u80fd\u53cd\u6620\u8fd9\u4e2a\u771f\u76f8\u3002<\/strong><\/li>\n\n\n\n
  • \u53ef\u52a0\u6027\uff1a\u591a\u65e5\u6536\u76ca\u53ef\u4ee5\u76f4\u63a5\u76f8\u52a0\uff0c\u65b9\u4fbf\u8ba1\u7b97\u3002<\/strong><\/li>\n\n\n\n
  • \u6807\u51c6\u5316\uff1a\u4e0d\u540c\u8d44\u4ea7\u3001\u4e0d\u540c\u65f6\u95f4\u5468\u671f\u90fd\u80fd\u653e\u5230\u4e00\u4e2a\u5c3a\u5ea6\u4e0a\u6bd4\u8f83\u3002<\/strong><\/li>\n<\/ul>\n\n\n\n
    \u5b9e\u64cd\u4ee3\u7801\uff1a\u6570\u636e\u5de5\u7a0b baseline<\/strong><\/h3>\n\n\n\n
    # \u5bfc\u5165\u5fc5\u8981\u7684\u5e93<\/code>import pandas as pd<\/code>import numpy as np<\/code>from openbb import obb<\/code>import matplotlib.pyplot as plt<\/code><\/code># \u7b2c\u4e00\u6b65\uff1a\u83b7\u53d6\u6807\u666e500\u6570\u636e\uff08SPY\uff09<\/code># \u7528OpenBB\u514d\u8d39\u6570\u636e\u6e90\uff0c\u6bd4\u5543Yahoo Finance\u7684\u63a5\u53e3\u7701\u4e8b\u591a\u4e86<\/code>data = obb.equity.price.historical(symbol=\"SPY\", provider=\"yfinance\")<\/code>df = data.to_df()<\/code><\/code># \u7b2c\u4e8c\u6b65\uff1a\u8ba1\u7b977\u65e5\u79fb\u52a8\u5e73\u5747\uff08\u5e73\u6ed1\u77ed\u671f\u566a\u97f3\uff09<\/code>df['ma_7'] = df['close'].rolling(window=7).mean()<\/code><\/code># \u7b2c\u4e09\u6b65\uff1a\u8ba1\u7b97\u5bf9\u6570\u6536\u76ca\u7387<\/code># log(\u4eca\u65e5\u4ef7\u683c\/\u6628\u65e5\u4ef7\u683c) * 100\uff0c\u53d8\u6210\u767e\u5206\u6bd4\u66f4\u76f4\u89c2<\/code>df['log_return'] = np.log(df['close'] \/ df['close'].shift(1)) * 100<\/code><\/code># \u5220\u9664\u7a7a\u503c\uff08\u524d7\u5929\u6ca1\u6709\u79fb\u52a8\u5e73\u5747\u503c\uff09<\/code>df.dropna(inplace=True)<\/code><\/code>print(df[['close', 'ma_7', 'log_return']].tail())<\/code><\/code><\/pre>\n\n\n\n
    \u4ee3\u7801\u89e3\u8bfb<\/strong>\uff1a<\/p>\n\n\n\n
      \n
    • rolling(7)\u50cf\u7ed9\u6570\u636e\u7a7f\u4ef6\"\u5e73\u6ed1\u8863\"\uff0c\u53bb\u6389\u77ed\u671f\u6742\u97f3\u3002<\/code><\/li>\n\n\n\n
    • np.log()\u662f\u6838\u5fc3\u8f6c\u6362\uff0c\u628a\u4ef7\u683c\u53d8\u6210\"\u6536\u76ca\u7387\u8bed\u8a00\"\u3002<\/code><\/li>\n\n\n\n
    • \u6700\u540e\u4e09\u884c\u6253\u5370\uff0c\u770b\u770b\u6211\u4eec\u7684\"\u539f\u6750\u6599\"\u957f\u5565\u6837\u3002<\/code><\/li>\n<\/ul>\n\n\n\n
      \u6b65\u9aa4<\/th>\u64cd\u4f5c<\/th>\u76ee\u7684<\/th>\u6ce8\u610f\u4e8b\u9879<\/th><\/tr><\/thead>
      \u6570\u636e\u83b7\u53d6<\/td>OpenBB API<\/td>\u514d\u8d39\u3001\u7a33\u5b9a\u3001\u65e0\u9700\u6ce8\u518c<\/td>\u68c0\u67e5\u6570\u636e\u5b8c\u6574\u6027<\/td><\/tr>
      \u5e73\u6ed1\u5904\u7406<\/td>7\u65e5\u79fb\u52a8\u5e73\u5747<\/td>\u8fc7\u6ee4\u566a\u97f3\uff0c\u4fdd\u7559\u8d8b\u52bf<\/td>\u7a97\u53e3\u671f\u53ef\u8c03\u6574<\/td><\/tr>
      \u6536\u76ca\u8f6c\u6362<\/td>\u5bf9\u6570\u6536\u76ca\u7387<\/td>\u5bf9\u79f0\u6027\u3001\u53ef\u6bd4\u6027<\/td>\u4e58\u4ee5100\u53d8\u767e\u5206\u6bd4<\/td><\/tr>
      \u7a7a\u503c\u6e05\u7406<\/td>dropna()<\/td>\u907f\u514d\u8ba1\u7b97\u9519\u8bef<\/td>\u68c0\u67e5\u5220\u9664\u6bd4\u4f8b<\/td><\/tr><\/tbody><\/table><\/figure>\n\n\n\n
      \u2462 \u4e09\u5927\u8bc6\u522b\u65b9\u6cd5\uff1a\u516b\u4ed9\u8fc7\u6d77\u5404\u663e\u795e\u901a<\/strong><\/h2>\n\n\n\n

      \u65b9\u6cd5\u4e00\uff1a\u9690\u9a6c\u5c14\u53ef\u592b\u6a21\u578b\uff08HMM\uff09\uff1a\u6700\u61c2”\u524d\u540e\u6587”\u7684\u6a21\u578b<\/strong><\/h3>\n\n\n\n
      \u6838\u5fc3\u601d\u60f3<\/strong>\uff1a\u5e02\u573a\u72b6\u6001\u5c31\u50cf\u5929\u6c14\uff0c\u867d\u7136\u770b\u4e0d\u89c1\uff0c\u4f46\u4f1a\u7559\u4e0b\u75d5\u8ff9\u3002\u4eca\u5929\u7684\u72b6\u6001\u4f1a\u5f71\u54cd\u660e\u5929\uff0c\u5c31\u50cf\u4e0b\u96e8\u4e4b\u540e\u5927\u6982\u7387\u8fd8\u662f\u9634\u5929\u3002<\/p>\n\n\n\n
      \u6a21\u578b\u7279\u70b9<\/strong>\uff1a<\/p>\n\n\n\n
        \n
      • \u5047\u8bbe\u5e02\u573a\u67093-4\u4e2a”\u9690\u85cf\u72b6\u6001”\uff1b<\/strong><\/li>\n\n\n\n
      • \u6bcf\u4e2a\u72b6\u6001\u4f1a\u4ea7\u751f\u4e0d\u540c\u7684\u6536\u76ca\u7387\u5206\u5e03\uff1b<\/strong><\/li>\n\n\n\n
      • \u80fd\u8ba1\u7b97\u72b6\u6001\u8f6c\u79fb\u6982\u7387\uff08\u6bd4\u5982\u4ece”\u5e73\u7a33”\u5230”\u66b4\u8e81”\u7684\u6982\u7387\uff09\u3002<\/strong><\/li>\n<\/ul>\n\n\n\n
        from hmmlearn.hmm import GaussianHMM<\/code># \u51c6\u5907\u6570\u636e\uff1a\u7528\u5bf9\u6570\u6536\u76ca\u7387\u4f5c\u4e3a\u89c2\u6d4b\u503c<\/code>X = df[['log_return']].values<\/code><\/code># \u521d\u59cb\u5316\u6a21\u578b\uff1a\u5047\u8bbe\u67093\u4e2a\u9690\u85cf\u72b6\u6001<\/code># n_components=3 \u5c31\u662f\u5047\u8bbe\u5e02\u573a\u67093\u79cd\u6027\u683c<\/code># covariance_type=\"full\" \u5141\u8bb8\u4e0d\u540c\u72b6\u6001\u6709\u4e0d\u540c\u7684\u6ce2\u52a8\u7387<\/code>model = GaussianHMM(n_components=3, covariance_type=\"full\", random_state=42)<\/code><\/code># \u8bad\u7ec3\u6a21\u578b<\/code>model.fit(X)<\/code><\/code># \u9884\u6d4b\u6bcf\u4e2a\u65f6\u95f4\u70b9\u6700\u53ef\u80fd\u7684\u72b6\u6001<\/code>hidden_states = model.predict(X)<\/code><\/code># \u628a\u9884\u6d4b\u7ed3\u679c\u52a0\u5230DataFrame\u91cc<\/code>df['hmm_state'] = hidden_states<\/code>print(f\"\u6a21\u578b\u53d1\u73b0\u76843\u79cd\u72b6\u6001\u53c2\u6570\uff1a\")<\/code>for i in range(model.n_components):<\/code>    print(f\"\u72b6\u6001{i}: \u5747\u503c={model.means_[i][0]:.2f}%, \u65b9\u5dee={np.sqrt(model.covars_[i][0][0]):.2f}%\")<\/code><\/code><\/pre>\n\n\n\n
        \"\"\/<\/figure>\n\n\n\n
        \u56fe2\uff1aHMM\u8bc6\u522b\u7684\u5e02\u573a\u72b6\u6001\u65f6\u95f4\u5e8f\u5217<\/em>\u6570\u636e\u6765\u6e90\uff1aSPY 2020-2025\u5e74\u65e5\u7ebf\uff0c\u8001\u4f59\u635e\u9c7c\u56e2\u961f\u56de\u6d4b<\/em><\/p>\n\n\n\n
        \u4f18\u52bf<\/strong>\uff1a\u61c2\u65f6\u5e8f\uff0c\u80fd\u6355\u6349\u72b6\u6001\u8f6c\u6362\u3002\u6ce8\u610f<\/strong>\uff1a\u9700\u8981\u8c03\u53c2\uff0cn_components<\/code>\u8bbe\u591a\u5c11\u5408\u9002\u5f97\u8bd5\u3002<\/p>\n\n\n\n
        \u65b9\u6cd5\u4e8c\uff1aK-Means\u805a\u7c7b\uff1a\u6700”\u76f4\u7537”\u7684\u5206\u7ec4\u65b9\u6cd5<\/strong><\/h3>\n\n\n\n
        \u6838\u5fc3\u601d\u60f3<\/strong>\uff1a\u4e0d\u770b\u65f6\u95f4\u5148\u540e\uff0c\u53ea\u770b\u8c01\u8ddf\u8c01\u957f\u5f97\u50cf\u3002\u628a\u6536\u76ca\u7387\u6570\u636e\u70b9\u5f53\u6210\u4e00\u7fa4\u5c0f\u670b\u53cb\uff0c\u8ba9\u76f8\u4f3c\u7684\u81ea\u5df1\u62b1\u56e2\u3002<\/p>\n\n\n\n
        \u6a21\u578b\u7279\u70b9<\/strong>\uff1a<\/p>\n\n\n\n