{"id":2305,"date":"2025-05-28T07:43:00","date_gmt":"2025-05-27T23:43:00","guid":{"rendered":"https:\/\/blog.laoyulaoyu.top\/?p=2305"},"modified":"2025-05-21T13:55:11","modified_gmt":"2025-05-21T05:55:11","slug":"%e6%9c%ba%e5%99%a8%e5%ad%a6%e4%b9%a0%e5%9c%a8%e5%b8%81%e5%9c%88%ef%bc%9a%e9%9a%8f%e6%9c%ba%e6%a3%ae%e6%9e%97%e6%a8%a1%e5%9e%8b%e5%ae%9e%e6%88%98%e9%a2%84%e6%b5%8b%e6%af%94%e7%89%b9%e5%b8%81%e6%98%8e","status":"publish","type":"post","link":"https:\/\/laoyulaoyu.com\/index.php\/2025\/05\/28\/%e6%9c%ba%e5%99%a8%e5%ad%a6%e4%b9%a0%e5%9c%a8%e5%b8%81%e5%9c%88%ef%bc%9a%e9%9a%8f%e6%9c%ba%e6%a3%ae%e6%9e%97%e6%a8%a1%e5%9e%8b%e5%ae%9e%e6%88%98%e9%a2%84%e6%b5%8b%e6%af%94%e7%89%b9%e5%b8%81%e6%98%8e\/","title":{"rendered":"\u673a\u5668\u5b66\u4e60\u5728\u5e01\u5708\uff1a\u968f\u673a\u68ee\u6797\u6a21\u578b\u5b9e\u6218\u9884\u6d4b\u6bd4\u7279\u5e01\u660e\u65e5\u6da8\u8dcc"},"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>\u6211\u7528Python\u548c\u673a\u5668\u5b66\u4e60\uff0c\u642d\u914d\u6280\u672f\u6307\u6807\uff0c\u6253\u9020\u4e86\u4e00\u4e2a\u9884\u6d4b\u6bd4\u7279\u5e01\u6da8\u8dcc\u7684\u5c0f\u5de5\u5177\u3002\u901a\u8fc7\u968f\u673a\u68ee\u6797\u6a21\u578b\uff0c\u5206\u6790\u5386\u53f2\u6570\u636e\uff0c\u5c1d\u8bd5\u5224\u65ad\u7b2c\u4e8c\u5929\u7684\u4ef7\u683c\u65b9\u5411\u3002<\/mark>\u867d\u7136\u9884\u6d4b\u4e0d\u662f\u767e\u5206\u767e\u51c6\u786e\uff0c\u4f46\u8fd9\u4e2a\u8fc7\u7a0b\u8ba9\u5927\u5bb6\u66f4\u4e86\u89e3\u5e02\u573a\u8d70\u52bf\u3002\u4ee5\u4e0b\u662f\u6211\u6574\u7406\u7684\u5b9e\u6218\u6559\u7a0b\uff0c\u5e0c\u671b\u5bf9\u4f60\u6709\u5e2e\u52a9\u3002<\/pre>\n<\/blockquote>\n\n\n\n
\u4e00\u3001\u4e3a\u5565\u8981\u7528AI\u9884\u6d4b\u6bd4\u7279\u5e01\uff1f<\/strong><\/h2>\n\n\n\n
\u6bd4\u7279\u5e01\u5e02\u573a\u5c31\u50cf\u8fc7\u5c71\u8f66\uff0c\u66b4\u6da8\u66b4\u8dcc\u5168\u9760\u731c\uff1f\u5e84\u5bb6\u63a7\u76d8\u3001\u6d88\u606f\u4e71\u98de\uff0c\u6563\u6237\u53ea\u80fd\u88ab\u5272\u97ed\u83dc\uff1f<\/p>\n\n\n\n
\u522b\u614c\uff01AI\u673a\u5668\u5b66\u4e60<\/strong>\u80fd\u5e2e\u4f60\u4ece\u5386\u53f2\u6570\u636e\u4e2d\u627e\u5230\u89c4\u5f8b\u3002\u6bd4\u5982\uff0c\u968f\u673a\u68ee\u6797\u6a21\u578b\u901a\u8fc7\u5206\u6790\u8fc7\u53bb\u7684\u4ef7\u683c\u3001\u6210\u4ea4\u91cf\u3001\u6280\u672f\u6307\u6807\uff0c\u9884\u6d4b\u672a\u6765\u6da8\u8dcc\uff0c\u867d\u7136\u7ed3\u679c\u660e\u786e\u53ef\u80fd\u8fd8\u4e0d\u5b8c\u7f8e\uff0c\u4f46\u7edd\u5bf9\u6bd4\u968f\u673a\u629b\u786c\u5e01\u9760\u8c31\u3002<\/p>\n\n\n\n
\u4e3a\u4ec0\u4e48\u9009\u968f\u673a\u68ee\u6797\uff1f<\/strong><\/p>\n\n\n\n
    \n
  • \u6838\u5fc3\u601d\u8def <\/strong>\uff1a\u901a\u8fc7\u5206\u6790\u5386\u53f2\u4ef7\u683c\u6ce2\u52a8\u89c4\u5f8b\uff0c\u8ba9AI\u81ea\u52a8\u8bc6\u522b”\u4e0a\u6da8\u4fe1\u53f7”\u548c”\u4e0b\u8dcc\u4fe1\u53f7”\u3002\u5c31\u50cf\u8001\u80a1\u6c11\u770bK\u7ebf\u56fe\uff0c\u6211\u4eec\u6559\u4f1a\u7535\u8111\u770b\u61c2\u8fd9\u4e9b\u6570\u5b57\u5bc6\u7801\u3002<\/li>\n\n\n\n
  • \u7b80\u5355\u9ad8\u6548<\/strong>\uff1a\u6bd4\u795e\u7ecf\u7f51\u7edc\u8bad\u7ec3\u5feb\uff0c\u5c0f\u767d\u4e5f\u80fd\u4e0a\u624b\u3002<\/li>\n\n\n\n
  • \u6297\u8fc7\u62df\u5408<\/strong>\uff1a\u901a\u8fc7\u591a\u68f5\u51b3\u7b56\u6811\u6295\u7968\uff0c\u907f\u514d\u201c\u778e\u8499\u201d\u3002<\/li>\n\n\n\n
  • \u7279\u5f81\u91cd\u8981\u6027\u5206\u6790<\/strong>\uff1a\u81ea\u52a8\u8bc6\u522b\u5173\u952e\u6307\u6807\uff08\u5982RSI\u3001\u5927\u6237\u6301\u4ed3\u6bd4\u4f8b\uff09\u3002<\/li>\n<\/ul>\n\n\n\n
    \u4e8c\u3001\u624b\u628a\u624b\u6559\u5b66<\/strong><\/h2>\n\n\n\n
    \u63a5\u4e0b\u6765\uff0c\u8ba9\u6211\u4eec\u8bd5\u8bd5\u7528\u201c\u4ee3\u7801+\u903b\u8f91+AI\u673a\u5668\u5b66\u4e60\u201d\u6765\u505a\u4e2a\u7406\u6027\u5206\u6790\uff1a\u9884\u6d4b\u6bd4\u7279\u5e01\u7b2c\u4e8c\u5929\u662f\u6da8\u662f\u8dcc\u3002<\/p>\n\n\n\n
    2.1 \u83b7\u53d6\u6570\u636e\uff1a\u4ece\u5386\u53f2\u4e2d\u627e\u7ebf\u7d22<\/strong><\/h3>\n\n\n\n
    \u6211\u4eec\u5148\u7528yfinance<\/code>\u5e93\u4e00\u952e\u4e0b\u8f7d\u6570\u636e\uff0c\u83b7\u53d6\u6bd4\u7279\u5e01\uff08BTC-USD\uff09\u8fc7\u53bb\u7684\u4ef7\u683c\u6570\u636e\u3002\u522b\u53bb\u722c\u4ea4\u6613\u6240\u4e86\uff01<\/p>\n\n\n\n
    import yfinance as yf\n# \u4e0b\u8f7d\u6bd4\u7279\u5e01\u6570\u636e\uff082010\u5e74\u81f3\u4eca\uff09\ndata = yf.download(\"BTC-USD\", start=\"2010-01-01\", end=\"2025-05-20\")\n# \u4fdd\u7559\u5173\u952e\u5b57\u6bb5\uff1a\u5f00\u76d8\u4ef7\u3001\u6536\u76d8\u4ef7\u3001\u6700\u9ad8\u4ef7\u3001\u6700\u4f4e\u4ef7\u3001\u6210\u4ea4\u91cf\ndata = data[['Open', 'Close', 'High', 'Low', 'Volume']]<\/code><\/pre>\n\n\n\n
    \u83b7\u5f97\u4e86\u5386\u53f2\u6536\u76d8\u4ef7\u6253\u5370\u51fa\u56fe\u6765\u540e\u5982\u4e0b\uff1a<\/p>\n\n\n\n
    \"\"<\/figure>\n\n\n\n

    2.2 \u7279\u5f81\u5de5\u7a0b\uff1a\u5e84\u5bb6\u90fd\u5728\u770b\u54ea\u4e9b\u6307\u6807\uff1f<\/strong><\/h3>\n\n\n\n
    \u6bd4\u7279\u5e01\u6da8\u8dcc\u4e0d\u662f\u7384\u5b66\uff0c\u800c\u662f\u6570\u636e\u548c\u89c4\u5f8b\u7684\u6e38\u620f<\/strong>\uff01<\/p>\n\n\n\n
    \u6211\u4eec\u521b\u5efa\u4e86\u5f88\u591a\u6280\u672f\u6307\u6807\uff0c\u6bd4\u5982\u6bcf\u65e5\u56de\u62a5\u7387\u3001\u4ef7\u683c\u6ce2\u52a8\u8303\u56f4\u3001\u79fb\u52a8\u5e73\u5747\u7ebf\u7b49\u3002\u8fd9\u4e9b\u6307\u6807\u5e2e\u52a9\u6a21\u578b\u7406\u89e3\u5e02\u573a\u8d8b\u52bf\u3002\u4ee5\u4e0b\u662f\u6bd4\u8f83\u91cd\u8981\u7684\u51e0\u4e2a\uff1a<\/p>\n\n\n\n
    \u7279\u5f81\u7c7b\u578b<\/th>\u793a\u4f8b\u6307\u6807<\/th>\u4f5c\u7528\u8bf4\u660e<\/th><\/tr><\/thead>
    \u4ef7\u683c\u6ce2\u52a8<\/td>\u65e5\u5185\u6ce2\u5e45\uff08High-Low\uff09<\/td>\u53cd\u6620\u5e02\u573a\u60c5\u7eea\u6fc0\u70c8\u7a0b\u5ea6<\/td><\/tr>
    \u6280\u672f\u6307\u6807<\/td>RSI\uff0814\u65e5\uff09\u3001MACD<\/td>\u5224\u65ad\u8d85\u4e70\u8d85\u5356<\/td><\/tr>
    \u5927\u6237\u52a8\u5411<\/td>\u5927\u6237\u591a\u7a7a\u8d26\u6237\u6bd4\u4f8b<\/td>\u8ddf\u5e84\u64cd\u4f5c<\/td><\/tr>
    \u6210\u4ea4\u91cf\u53d8\u5316<\/td>5\u65e5\u6210\u4ea4\u91cf\u5747\u503c<\/td>\u8bc6\u522b\u4e3b\u529b\u8d44\u91d1\u8fdb\u51fa<\/td><\/tr>
    \u65f6\u95f4\u5468\u671f<\/td>\u661f\u671f\u51e0\u3001\u6708\u4efd<\/td>\u6355\u6349\u5468\u671f\u6027\u89c4\u5f8b<\/td><\/tr><\/tbody><\/table><\/figure>\n\n\n\n
    \u4ee3\u7801\u793a\u4f8b\uff08\u6dfb\u52a0RSI\u548cMACD\u6307\u6807\uff09<\/strong>\uff1a<\/p>\n\n\n\n
    from ta.momentum import RSIIndicator\nfrom ta.trend import MACD\n# \u8ba1\u7b9714\u65e5RSI\ndata['RSI'] = RSIIndicator(data['Close'], window=14).rsi()\n# \u8ba1\u7b97MACD\nmacd = MACD(data['Close'])\ndata['MACD'] = macd.macd()\ndata['MACD_Signal'] = macd.macd_signal()<\/code><\/pre>\n\n\n\n
    2.3 \u8bad\u7ec3\u6a21\u578b\uff1a\u968f\u673a\u68ee\u6797\u5b9e\u6218<\/strong><\/h3>\n\n\n\n
    \u7528\u968f\u673a\u68ee\u6797\u7b97\u6cd5\u8bad\u7ec3\u6a21\u578b\uff0c\u8fd9\u662f\u4e00\u79cd\u901a\u8fc7\u591a\u4e2a\u51b3\u7b56\u6811\u6295\u7968\u6765\u9884\u6d4b\u7684AI\u65b9\u6cd5\u3002<\/p>\n\n\n\n
    from sklearn.ensemble import RandomForestClassifier\nfrom sklearn.model_selection import train_test_split\n\n# \u76ee\u6807\u53d8\u91cf\uff1a\u6b21\u65e5\u6da8\u8dcc\uff081=\u6da8\uff0c0=\u8dcc\uff09\ndata['Target'] = (data['Close'].shift(-1) > data['Close']).astype(int)\ndata.dropna(inplace=True)\n\n# \u5212\u5206\u8bad\u7ec3\u96c6\u548c\u6d4b\u8bd5\u96c6\uff08\u6309\u65f6\u95f4\u987a\u5e8f\uff0c\u7981\u6b62\u6253\u4e71\uff01\uff09\nX = data.drop(['Target'], axis=1)\ny = data['Target']\nX_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.2, shuffle=False)\n\n# \u8bad\u7ec3\u968f\u673a\u68ee\u6797\uff08\u91cd\u70b9\u8c03\u53c2\uff1an_estimators=200, max_depth=10\uff09\nmodel = RandomForestClassifier(n_estimators=200, max_depth=10, class_weight='balanced')\nmodel.fit(X_train, y_train)<\/code><\/pre>\n\n\n\n
    \u6211\u4eec\u8fd8\u7ed8\u5236\u4e86\u524d 10 \u4e2a\u9009\u5b9a\u7279\u5f81\u7684 F \u5206\u6570\uff0c\u4ee5\u4e86\u89e3\u54ea\u4e9b\u7279\u5f81\u5bf9\u76ee\u6807\u7684\u8d21\u732e\u6700\u5927\u3002<\/p>\n\n\n\n
    \"\"<\/figure>\n\n\n\n
    \u907f\u5751\u6307\u5357<\/strong>\uff1a<\/p>\n\n\n\n
      \n
    • \u6570\u636e\u5fc5\u987b\u6309\u65f6\u95f4\u987a\u5e8f\u5206\u5272\uff0c\u5426\u5219\u4f1a\u201c\u672a\u6765\u4fe1\u606f\u6cc4\u9732\u201d\uff01<\/li>\n\n\n\n
    • \u7c7b\u522b\u4e0d\u5e73\u8861\u65f6\u7528class_weight='balanced'<\/code>\uff0c\u9632\u6b62\u6a21\u578b\u8eba\u5e73\u3002<\/li>\n<\/ul>\n\n\n\n
      2.4 \u7ed3\u679c\u5206\u6790\uff1a\u8bc6\u7834\u5e84\u5bb6\u9677\u9631<\/strong><\/h3>\n\n\n\n
      # \u9884\u6d4b\u6d4b\u8bd5\u96c6\ny_pred = model.predict(X_test)\n\n# \u6253\u5370\u51c6\u786e\u7387\u3001\u53ec\u56de\u7387\nfrom sklearn.metrics import classification_report\nprint(classification_report(y_test, y_pred))<\/code><\/pre>\n\n\n\n
      \u8f93\u51fa\u793a\u4f8b<\/strong>\uff1a<\/p>\n\n\n\n
                   precision    recall  f1-score   support

                 0       0.51      0.78      0.62       380
                 1       0.57      0.28      0.38       393

          accuracy                           0.53       773
         macro avg       0.54      0.53      0.50       773
      weighted avg       0.54      0.53      0.50       773<\/pre>\n\n\n\n
      \u89e3\u8bfb<\/strong>\uff1a\u4ece\u7ed3\u679c\u770b\uff0c\u6a21\u578b\u7684\u603b\u4f53\u51c6\u786e\u7387\u4e3a53%\uff0c\u6bd4\u968f\u673a\u731c\uff0850%\uff09\u7a0d\u597d\u70b9\u3002\u8be6\u7ec6\u770b\uff0c\u9884\u6d4b\u4ef7\u683c\u4e0d\u6da8\uff08\u7c7b\u522b0\uff09\u65f6\uff0c\u53ec\u56de\u738778%\uff0c\u8868\u73b0\u4e0d\u9519\uff1b\u4f46\u9884\u6d4b\u4ef7\u683c\u6da8\uff08\u7c7b\u522b1\uff09\u65f6\uff0c\u53ec\u56de\u7387\u4ec528%\uff0c\u6548\u679c\u8f83\u5dee\u3002\u8fd9\u8bf4\u660e\u6a21\u578b\u66f4\u64c5\u957f\u53d1\u73b0\u4ef7\u683c\u4e0d\u6da8\u7684\u60c5\u51b5\u3002<\/p>\n\n\n\n
      \u63a5\u4e0b\u6765\u8fdb\u884c\u9884\u6d4b\u5e02\u573a\u8d70\u52bf\u7684\u56fe\u5f62\u53ef\u89c6\u5316\u3002<\/strong><\/p>\n\n\n\n
      plt.figure(figsize=( 14 , 6 )) \nplt.plot(data.index[- len (y_test):], data[ \"\u6536\u76d8\u4ef7\" ][- len (y_test):], label= '\u6536\u76d8\u4ef7' ) \nplt.plot(data.index[- len (y_test):][y_pred == 1 ], data[ \"\u6536\u76d8\u4ef7\" ][- len (y_test):][y_pred == 1 ], '^' , markersize= 10 , color= 'g' , label= '\u9884\u6d4b\u4e0a\u6da8' ) \nplt.plot(data.index[- len (y_test):][y_pred == 0 ], data[ \"\u6536\u76d8\u4ef7\" ][- len (y_test):][y_pred == 0 ], 'v' , markersize= 10 , color= 'r' , label= '\u9884\u6d4b\u4e0b\u8dcc' ) \nplt.title( \u201c\u9884\u6d4b\u6bd4\u7279\u5e01\u65b9\u5411\u4e0e\u6536\u76d8\u4ef7\u201d\uff09\nplt.legend\uff08\uff09\nplt.savefig\uff08\u201cpredicted_market_direction.png\u201d\uff09\nplt.show\uff08\uff09<\/code><\/pre>\n\n\n\n
      \u4e0b\u56fe\u8986\u76d6\u4e86\u6bd4\u7279\u5e01\u6536\u76d8\u4ef7\u7684\u6a21\u578b\u9884\u6d4b\uff1a<\/p>\n\n\n\n
      \"\"<\/figure>\n\n\n\n
        \n
      • \u7eff\u8272\u4e09\u89d2\u5f62\u8868\u793a\u6a21\u578b\u9884\u6d4b\u4e0a\u6da8\u7684\u65e5\u671f\u3002<\/li>\n\n\n\n
      • \u7ea2\u8272\u4e09\u89d2\u5f62\u8868\u793a\u9884\u6d4b\u7684\u5411\u4e0b\u8fd0\u52a8\u3002<\/li>\n<\/ul>\n\n\n\n
        \u6211\u4eec\u8fd8\u901a\u8fc7\u6df7\u6dc6\u77e9\u9635\u5feb\u901f\u8bc6\u522b\u6709\u591a\u5c11\u9884\u6d4b\u662f\u6b63\u786e\u548c\u9519\u8bef\u7684\u3002\u5b83\u533a\u5206\u4e86\uff1a<\/p>\n\n\n\n
        cm = confusion_matrix(y_test, y_pred) \nsns.heatmap(cm, annot= True , fmt= 'd' , cmap= 'Blues' ) \nplt.xlabel( '\u9884\u6d4b' ) \nplt.ylabel( '\u5b9e\u9645' ) \nplt.title( '\u6df7\u6dc6\u77e9\u9635' ) \nplt.savefig( \"confusion_matrix.png\" ) \nplt.show()<\/code><\/pre>\n\n\n\n
        \u6df7\u6dc6\u77e9\u9635\u53ef\u4ee5\u5feb\u901f\u663e\u793a\u6709\u591a\u5c11\u9884\u6d4b\u662f\u6b63\u786e\u548c\u9519\u8bef\u7684\u3002\u5b83\u533a\u5206\u4e86\uff1a<\/p>\n\n\n\n
        \"\"<\/figure>\n\n\n\n