\u76ee\u5f55<\/p> \n
- \n
- \u65f6\u95f4\u5e8f\u5217\u6df1\u5ea6\u5b66\u4e60\uff1a\u72b6\u6001 LSTM \u6a21\u578b\u9884\u6d4b\u592a\u9633\u9ed1\u5b50<\/a>\n
- \n
- \u6559\u7a0b\u6982\u89c8<\/a><\/li> \n
- \u5546\u4e1a\u5e94\u7528<\/a><\/li> \n
- \u957f\u77ed\u671f\u8bb0\u5fc6\uff08LSTM\uff09\u6a21\u578b<\/a><\/li> \n
- \u592a\u9633\u9ed1\u5b50\u6570\u636e\u96c6<\/a><\/li> \n
- \u6784\u5efa LSTM \u6a21\u578b\u9884\u6d4b\u592a\u9633\u9ed1\u5b50<\/a>\n
- \n
- 1 \u82e5\u5e72\u76f8\u5173\u5305<\/a><\/li> \n
- 2 \u6570\u636e<\/a><\/li> \n
- 3 \u63a2\u7d22\u6027\u6570\u636e\u5206\u6790<\/a><\/li> \n
- 4 \u56de\u6d4b\uff1a\u65f6\u95f4\u5e8f\u5217\u4ea4\u53c9\u9a8c\u8bc1<\/a><\/li> \n
- 5 \u7528 Keras \u6784\u5efa\u72b6\u6001 LSTM \u6a21\u578b<\/a><\/li> \n <\/ul><\/li> \n
- \u7ed3\u8bba<\/a><\/li> \n <\/ul><\/li> \n <\/ul> \n <\/div> \n <\/div> \n
\u65f6\u95f4\u5e8f\u5217\u6df1\u5ea6\u5b66\u4e60\uff1a\u72b6\u6001 LSTM \u6a21\u578b\u9884\u6d4b\u592a\u9633\u9ed1\u5b50<\/h1> \n
\n
\u672c\u6587\u7ffb\u8bd1\u81ea\u300aTime Series Deep Learning: Forecasting Sunspots With Keras Stateful Lstm In R\u300b<\/p> \n
\u539f\u6587\u94fe\u63a5<\/a><\/p> \n <\/blockquote> \n
<\/p> \n
\u7531\u4e8e\u6570\u636e\u79d1\u5b66\u673a\u5668\u5b66\u4e60\u548c\u6df1\u5ea6\u5b66\u4e60\u7684\u53d1\u5c55\uff0c\u65f6\u95f4\u5e8f\u5217\u9884\u6d4b\u5728\u9884\u6d4b\u51c6\u786e\u6027\u65b9\u9762\u53d6\u5f97\u4e86\u663e\u7740\u8fdb\u5c55\u3002<\/strong>\u968f\u7740\u8fd9\u4e9b ML\/DL \u5de5\u5177\u7684\u53d1\u5c55\uff0c\u4f01\u4e1a\u548c\u91d1\u878d\u673a\u6784\u73b0\u5728\u53ef\u4ee5\u901a\u8fc7\u5e94\u7528\u8fd9\u4e9b\u65b0\u6280\u672f\u6765\u89e3\u51b3\u65e7\u95ee\u9898\uff0c\u4ece\u800c\u66f4\u597d\u5730\u8fdb\u884c\u9884\u6d4b\u3002\u5728\u672c\u6587\u4e2d\uff0c\u6211\u4eec\u5c55\u793a\u4e86\u4f7f\u7528\u79f0\u4e3a LSTM\uff08\u957f\u77ed\u671f\u8bb0\u5fc6\uff09<\/a>\u7684\u7279\u6b8a\u7c7b\u578b\u6df1\u5ea6\u5b66\u4e60\u6a21\u578b\uff0c\u8be5\u6a21\u578b\u5bf9\u6d89\u53ca\u81ea\u76f8\u5173\u6027\u7684\u5e8f\u5217\u9884\u6d4b\u95ee\u9898\u5f88\u6709\u7528\u3002\u6211\u4eec\u5206\u6790\u4e86\u4e00\u4e2a\u540d\u4e3a\u201c\u592a\u9633\u9ed1\u5b50\u201d<\/a>\u7684\u8457\u540d\u5386\u53f2\u6570\u636e\u96c6\uff08\u592a\u9633\u9ed1\u5b50<\/a>\u662f\u6307\u592a\u9633\u8868\u9762\u5f62\u6210\u9ed1\u70b9\u7684\u592a\u9633\u73b0\u8c61)\u3002\u6211\u4eec\u5c06\u5c55\u793a\u5982\u4f55\u4f7f\u7528 LSTM \u6a21\u578b\u9884\u6d4b\u672a\u6765 10 \u5e74\u7684\u592a\u9633\u9ed1\u5b50\u6570\u91cf\u3002<\/p> \n
\u6559\u7a0b\u6982\u89c8<\/h2> \n
\u6b64\u4ee3\u7801\u6559\u7a0b\u5bf9\u5e94\u4e8e 2018 \u5e74 4 \u6708 19 \u65e5\u661f\u671f\u56db\u5411 SP Global<\/a> \u63d0\u4f9b\u7684 Time Series Deep Learning \u6f14\u793a\u6587\u7a3f<\/a>\u3002\u53ef\u4ee5\u4e0b\u8f7d\u8865\u5145\u672c\u6587\u7684\u5e7b\u706f\u7247\u3002<\/p> \n
\u8fd9\u662f\u4e00\u4e2a\u6d89\u53ca\u65f6\u95f4\u5e8f\u5217\u6df1\u5ea6\u5b66\u4e60<\/strong>\u548c\u5176\u4ed6\u590d\u6742\u673a\u5668\u5b66\u4e60\u4e3b\u9898\uff08\u5982\u56de\u6d4b\u4ea4\u53c9\u9a8c\u8bc1\uff09\u7684\u9ad8\u7ea7\u6559\u7a0b<\/strong>\u3002\u5982\u679c\u60f3\u8981\u4e86\u89e3 R \u4e2d\u7684 Keras<\/strong>\uff0c\u8bf7\u67e5\u770b\uff1aCustomer Analytics: Using Deep Learning With Keras To Predict Customer Churn<\/a>\u3002<\/p> \n
\u672c\u6559\u7a0b\u4e2d\uff0c\u4f60\u5c06\u4f1a\u5b66\u5230\uff1a<\/p> \n
- \n
- \u7528
keras<\/code> \u5305\u5f00\u53d1\u4e00\u4e2a\u72b6\u6001 LSTM \u6a21\u578b<\/strong>\uff0c\u8be5 R \u5305\u5c06 R TensorFlow<\/a> \u4f5c\u4e3a\u540e\u7aef\u3002<\/li> \n
- \u5c06\u72b6\u6001 LSTM \u6a21\u578b\u5e94\u7528\u5230\u8457\u540d\u7684\u592a\u9633\u9ed1\u5b50<\/strong>\u6570\u636e\u96c6\u4e0a\u3002<\/li> \n
- \u501f\u52a9
rsample<\/code> \u5305\u5728\u521d\u59cb\u62bd\u6837\u4e0a\u6eda\u52a8\u9884\u6d4b<\/a>\uff0c\u5b9e\u73b0\u65f6\u95f4\u5e8f\u5217\u7684\u4ea4\u53c9\u68c0\u9a8c<\/strong>\u3002<\/li> \n
- \u501f\u52a9
ggplot2<\/code> \u548c
cowplot<\/code> \u53ef\u89c6\u5316\u56de\u6d4b\u548c\u9884\u6d4b\u7ed3\u679c\u3002<\/li> \n
- \u901a\u8fc7\u81ea\u76f8\u5173\u51fd\u6570\uff08Autocorrelation Function\uff0cACF\uff09\u56fe<\/strong>\u8bc4\u4f30\u65f6\u95f4\u5e8f\u5217\u6570\u636e\u662f\u5426\u9002\u5408\u5e94\u7528 LSTM \u6a21\u578b\u3002<\/li> \n <\/ul> \n
\u672c\u6587\u7684\u6700\u7ec8\u7ed3\u679c\u662f\u4e00\u4e2a\u9ad8\u6027\u80fd\u7684\u6df1\u5ea6\u5b66\u4e60\u7b97\u6cd5<\/strong>\uff0c\u5728\u9884\u6d4b\u672a\u6765 10 \u5e74\u592a\u9633\u9ed1\u5b50\u6570\u91cf\u65b9\u9762\u8868\u73b0\u975e\u5e38\u51fa\u8272\uff01\u8fd9\u662f\u56de\u6d4b\u540e\u72b6\u6001 LSTM \u6a21\u578b<\/strong>\u7684\u7ed3\u679c\u3002<\/p> \n
<\/p> \n
\u5546\u4e1a\u5e94\u7528<\/h2> \n
\u65f6\u95f4\u5e8f\u5217\u9884\u6d4b\u5bf9\u8425\u6536\u548c\u5229\u6da6\u6709\u663e\u7740\u5f71\u54cd\u3002\u5728\u5546\u4e1a\u65b9\u9762\uff0c\u6211\u4eec\u53ef\u80fd\u6709\u5174\u8da3\u9884\u6d4b\u6bcf\u6708\u3001\u6bcf\u5b63\u5ea6\u6216\u6bcf\u5e74\u7684\u54ea\u4e00\u5929\u4f1a\u53d1\u751f\u5927\u989d\u652f\u51fa\uff0c\u6216\u8005\u6211\u4eec\u53ef\u80fd\u6709\u5174\u8da3\u4e86\u89e3\u6d88\u8d39\u8005\u7269\u4ef7\u6307\u6570\uff08CPI\uff09\u5728\u672a\u6765\u516d\u5e74\u4e2a\u6708\u5982\u4f55\u53d8\u5316\u3002\u8fd9\u4e9b\u90fd\u662f\u5728\u5fae\u89c2\u7ecf\u6d4e\u548c\u5b8f\u89c2\u7ecf\u6d4e\u5c42\u9762\u5f71\u54cd\u5546\u4e1a\u7ec4\u7ec7\u7684\u5e38\u89c1\u95ee\u9898\u3002\u867d\u7136\u672c\u6559\u7a0b\u4e2d\u4f7f\u7528\u7684\u6570\u636e\u96c6\u4e0d\u662f\u201c\u5546\u4e1a\u201d\u6570\u636e\u96c6\uff0c\u4f46\u5b83\u663e\u793a\u4e86\u5de5\u5177-\u95ee\u9898\u5339\u914d<\/strong>\u7684\u5f3a\u5927\u529b\u91cf\uff0c\u610f\u5473\u7740\u4f7f\u7528\u6b63\u786e\u7684\u5de5\u5177\u8fdb\u884c\u5de5\u4f5c\u53ef\u4ee5\u5927\u5927\u63d0\u9ad8\u51c6\u786e\u6027\u3002\u6700\u7ec8\u7684\u7ed3\u679c\u662f\u9884\u6d4b\u51c6\u786e\u6027\u7684\u63d0\u9ad8\u5c06\u5bf9\u8425\u6536\u548c\u5229\u6da6\u5e26\u6765\u53ef\u91cf\u5316\u7684\u63d0\u5347\u3002<\/p> \n
\u957f\u77ed\u671f\u8bb0\u5fc6\uff08LSTM\uff09\u6a21\u578b<\/h2> \n
\u957f\u77ed\u671f\u8bb0\u5fc6\uff08LSTM\uff09\u6a21\u578b<\/strong>\u662f\u4e00\u79cd\u5f3a\u5927\u7684\u9012\u5f52\u795e\u7ecf\u7f51\u7edc\uff08RNN\uff09\u3002\u535a\u6587\u300aUnderstanding LSTM Networks<\/a>\u300b\uff08\u7ffb\u8bd1\u7248<\/a>)\u4ee5\u7b80\u5355\u6613\u61c2\u7684\u65b9\u5f0f\u89e3\u91ca\u4e86\u6a21\u578b\u7684\u590d\u6742\u6027\u673a\u5236\u3002\u4e0b\u9762\u662f\u63cf\u8ff0 LSTM \u5185\u90e8\u5355\u5143\u67b6\u6784\u7684\u793a\u610f\u56fe\uff0c\u9664\u77ed\u671f\u72b6\u6001\u4e4b\u5916\uff0c\u8be5\u67b6\u6784\u4f7f\u5176\u80fd\u591f\u4fdd\u6301\u957f\u671f\u72b6\u6001\uff0c\u800c\u8fd9\u662f\u4f20\u7edf\u7684 RNN \u5904\u7406\u8d77\u6765\u6709\u56f0\u96be\u7684\uff1a<\/p> \n
<\/p> \n
\u6765\u6e90\uff1aUnderstanding LSTM Networks<\/a><\/p> \n
LSTM \u6a21\u578b\u5728\u9884\u6d4b\u5177\u6709\u81ea\u76f8\u5173\u6027\uff08\u65f6\u95f4\u5e8f\u5217\u548c\u6ede\u540e\u9879\u4e4b\u95f4\u5b58\u5728\u76f8\u5173\u6027\uff09\u7684\u65f6\u95f4\u5e8f\u5217\u65f6\u975e\u5e38\u6709\u7528\uff0c\u56e0\u4e3a\u6a21\u578b\u80fd\u591f\u4fdd\u6301\u72b6\u6001\u5e76\u8bc6\u522b\u65f6\u95f4\u5e8f\u5217\u4e0a\u7684\u6a21\u5f0f\u3002\u5728\u6bcf\u6b21\u5904\u7406\u8fc7\u7a0b\u4e2d\uff0c\u9012\u5f52\u67b6\u6784\u80fd\u4f7f\u72b6\u6001\u5728\u66f4\u65b0\u6743\u91cd\u65f6\u4fdd\u6301\u6216\u8005\u4f20\u9012\u4e0b\u53bb\u3002\u6b64\u5916\uff0cLSTM \u6a21\u578b\u7684\u5355\u5143\u67b6\u6784\u5728\u77ed\u671f\u6301\u4e45\u5316\u7684\u57fa\u7840\u4e0a\u5b9e\u73b0\u4e86\u957f\u671f\u6301\u4e45\u5316\uff0c\u8fdb\u800c\u5f3a\u5316\u4e86 RNN\uff0c\u8fd9\u4e00\u70b9\u975e\u5e38\u5438\u5f15\u4eba\uff01<\/p> \n
\u5728 Keras \u4e2d\uff0cLSTM \u6a21\u578b\u53ef\u4ee5\u6709\u201c\u72b6\u6001\u201d\u6a21\u5f0f\uff0cKeras \u6587\u6863\u4e2d\u8fd9\u6837\u89e3\u91ca\uff1a<\/p> \n
\n
\u7d22\u5f15 i \u5904\u6bcf\u4e2a\u6837\u672c\u7684\u6700\u540e\u72b6\u6001\u5c06\u88ab\u7528\u4f5c\u4e0b\u4e00\u6b21\u6279\u5904\u7406\u4e2d\u7d22\u5f15 i \u5904\u6837\u672c\u7684\u521d\u59cb\u72b6\u6001<\/p> \n <\/blockquote> \n
\u5728\u6b63\u5e38\uff08\u6216\u201c\u65e0\u72b6\u6001\u201d\uff09\u6a21\u5f0f\u4e0b\uff0cKeras \u5bf9\u6837\u672c\u91cd\u65b0\u6d17\u724c\uff0c\u65f6\u95f4\u5e8f\u5217\u4e0e\u5176\u6ede\u540e\u9879\u4e4b\u95f4\u7684\u4f9d\u8d56\u5173\u7cfb\u4e22\u5931\u3002\u4f46\u662f\uff0c\u5728\u201c\u72b6\u6001\u201d\u6a21\u5f0f\u4e0b\u8fd0\u884c\u65f6\uff0c\u6211\u4eec\u901a\u5e38\u53ef\u4ee5\u901a\u8fc7\u5229\u7528\u65f6\u95f4\u5e8f\u5217\u4e2d\u5b58\u5728\u7684\u81ea\u76f8\u5173\u6027\u6765\u83b7\u5f97\u9ad8\u8d28\u91cf\u7684\u9884\u6d4b\u7ed3\u679c<\/strong>\u3002<\/p> \n
\u5728\u5b8c\u6210\u672c\u6559\u7a0b\u65f6\uff0c\u6211\u4eec\u4f1a\u8fdb\u4e00\u6b65\u89e3\u91ca\u3002\u5c31\u76ee\u524d\u800c\u8a00\uff0c\u53ef\u4ee5\u8ba4\u4e3a LSTM \u6a21\u578b\u5bf9\u6d89\u53ca\u81ea\u76f8\u5173\u6027\u7684\u65f6\u95f4\u5e8f\u5217\u95ee\u9898\u53ef\u80fd\u975e\u5e38\u6709\u7528\uff0c\u800c\u4e14 Keras \u6709\u80fd\u529b\u521b\u5efa\u5b8c\u7f8e\u7684\u65f6\u95f4\u5e8f\u5217\u5efa\u6a21\u5de5\u5177\u2014\u2014\u72b6\u6001 LSTM \u6a21\u578b\u3002<\/p> \n
\u592a\u9633\u9ed1\u5b50\u6570\u636e\u96c6<\/h2> \n
\u592a\u9633\u9ed1\u5b50<\/a>\u662f\u968f R \u53d1\u5e03\u7684\u8457\u540d\u6570\u636e\u96c6\uff08\u53c2\u89c1
datasets<\/code> \u5305\uff09\u3002\u6570\u636e\u96c6\u8ddf\u8e2a\u8bb0\u5f55\u592a\u9633\u9ed1\u5b50\uff0c\u5373\u592a\u9633\u8868\u9762\u51fa\u73b0\u9ed1\u70b9\u7684\u4e8b\u4ef6\u3002\u8fd9\u662f\u6765\u81ea NASA \u7684\u4e00\u5f20\u7167\u7247\uff0c\u663e\u793a\u4e86\u592a\u9633\u9ed1\u5b50\u73b0\u8c61\u3002\u76f8\u5f53\u9177\uff01<\/p> \n
<\/p> \n
\u6765\u6e90\uff1aNASA<\/a><\/p> \n
\u672c\u6559\u7a0b\u6240\u7528\u7684\u6570\u636e\u96c6\u79f0\u4e3a
sunspots.month<\/code>\uff0c\u5305\u542b\u4e86 265\uff081749 \uff5e 2013\uff09\u5e74\u95f4\u6bcf\u6708\u592a\u9633\u9ed1\u5b50\u6570\u91cf\u7684\u6708\u5ea6\u6570\u636e\u3002<\/p> \n
<\/p> \n
\u6784\u5efa LSTM \u6a21\u578b\u9884\u6d4b\u592a\u9633\u9ed1\u5b50<\/h2> \n
\u8ba9\u6211\u4eec\u5f00\u52a8\u8d77\u6765\uff0c\u9884\u6d4b\u592a\u9633\u9ed1\u5b50\u3002\u8fd9\u662f\u6211\u4eec\u7684\u76ee\u6807\uff1a<\/p> \n
\u76ee\u6807<\/strong>\uff1a\u4f7f\u7528 LSTM \u6a21\u578b\u9884\u6d4b\u672a\u6765 10 \u5e74<\/strong>\u7684\u592a\u9633\u9ed1\u5b50\u6570\u91cf\u3002<\/p> \n
1 \u82e5\u5e72\u76f8\u5173\u5305<\/h3> \n
\u4ee5\u4e0b\u662f\u672c\u6559\u7a0b\u6240\u9700\u7684\u5305\uff0c\u6240\u6709\u8fd9\u4e9b\u5305\u90fd\u53ef\u4ee5\u5728 CRAN \u4e0a\u627e\u5230\u3002\u5982\u679c\u4f60\u5c1a\u672a\u5b89\u88c5\u8fd9\u4e9b\u5305\uff0c\u53ef\u4ee5\u4f7f\u7528
install.packages()<\/code> \u8fdb\u884c\u5b89\u88c5\u3002\u6ce8\u610f\uff1a\u5728\u7ee7\u7eed\u4f7f\u7528\u6b64\u4ee3\u7801\u6559\u7a0b\u4e4b\u524d\uff0c\u8bf7\u786e\u4fdd\u66f4\u65b0\u6240\u6709\u5305\uff0c\u56e0\u4e3a\u8fd9\u4e9b\u5305\u7684\u5148\u524d\u7248\u672c\u53ef\u80fd\u4e0e\u6240\u7528\u4ee3\u7801\u4e0d\u517c\u5bb9\u3002<\/strong><\/p> \n
# Core Tidyverse\nlibrary(tidyverse)\nlibrary(glue)\nlibrary(forcats)\n\n# Time Series\nlibrary(timetk)\nlibrary(tidyquant)\nlibrary(tibbletime)\n\n# Visualization\nlibrary(cowplot)\n\n# Preprocessing\nlibrary(recipes)\n\n# Sampling \/ Accuracy\nlibrary(rsample)\nlibrary(yardstick)\n\n# Modeling\nlibrary(keras)<\/code><\/pre> \n
\u5982\u679c\u4f60\u4e4b\u524d\u6ca1\u6709\u5728 R \u4e2d\u8fd0\u884c\u8fc7 Keras\uff0c\u4f60\u9700\u8981\u7528
install_keras()<\/code> \u51fd\u6570\u5b89\u88c5 Keras\u3002<\/p> \n
# Install Keras if you have not installed before\ninstall_keras()<\/code><\/pre> \n
2 \u6570\u636e<\/h3> \n
\u6570\u636e\u96c6
sunspot.month<\/code> \u968f R \u4e00\u8d77\u53d1\u5e03\uff0c\u53ef\u4ee5\u8f7b\u6613\u83b7\u5f97\u3002\u5b83\u662f\u4e00\u4e2a
ts<\/code> \u7c7b\u5bf9\u8c61\uff08\u975e tidy \u7c7b\uff09\uff0c\u6240\u4ee5\u6211\u4eec\u5c06\u4f7f\u7528
timetk<\/code> \u4e2d\u7684
tk_tbl()<\/code> \u51fd\u6570\u8f6c\u6362\u4e3a tidy \u6570\u636e\u96c6\u3002\u6211\u4eec\u4f7f\u7528\u8fd9\u4e2a\u51fd\u6570\u800c\u4e0d\u662f\u6765\u81ea
tibble<\/code> \u7684
as.tibble()<\/code>\uff0c\u7528\u6765\u81ea\u52a8\u5c06\u65f6\u95f4\u5e8f\u5217\u7d22\u5f15\u4fdd\u5b58\u4e3a
zoo<\/code>
yearmon<\/code> \u7d22\u5f15\u3002\u6700\u540e\uff0c\u6211\u4eec\u5c06\u4f7f\u7528
lubridate::as_date()<\/code>\uff08\u4f7f\u7528
tidyquant<\/code> \u65f6\u52a0\u8f7d\uff09\u5c06
zoo<\/code> \u7d22\u5f15\u8f6c\u6362\u4e3a\u65e5\u671f\uff0c\u7136\u540e\u8f6c\u6362\u4e3a
tbl_time<\/code> \u5bf9\u8c61\u4ee5\u4f7f\u65f6\u95f4\u5e8f\u5217\u64cd\u4f5c\u8d77\u6765\u66f4\u5bb9\u6613\u3002<\/p> \n
sun_spots <- datasets::sunspot.month %>%\n tk_tbl() %>%\n mutate(index = as_date(index)) %>%\n as_tbl_time(index = index)\n\nsun_spots<\/code><\/pre> \n
## # A time tibble: 3,177 x 2\n## # Index: index\n## index value\n## <date> <dbl>\n## 1 1749-01-01 58.0\n## 2 1749-02-01 62.6\n## 3 1749-03-01 70.0\n## 4 1749-04-","orderid":"0","title":"\u65f6\u95f4\u5e8f\u5217\u6df1\u5ea6\u5b66\u4e60\uff1a\u72b6\u6001 LSTM \u6a21\u578b\u9884\u6d4b\u592a\u9633\u9ed1\u5b50(\u4e00)","smalltitle":"","mid":"0","fname":"R\u8bed\u8a00","special_id":"0","bak_id":"0","info":"0","hits":"667","pages":"10","comments":"0","posttime":"2019-09-03 02:41:30","list":"1567449690","username":"admin","author":"","copyfrom":"","copyfromurl":"","titlecolor":"","fonttype":"0","titleicon":"0","picurl":"https:\/\/www.cppentry.com\/upload_files\/","ispic":"0","yz":"1","yzer":"","yztime":"0","levels":"0","levelstime":"0","keywords":"\u65f6\u95f4\u5e8f\u5217<\/A> \u6df1\u5ea6<\/A> \u5b66\u4e60<\/A> \u72b6\u6001<\/A> LSTM<\/A> \u6a21\u578b<\/A> \u9884\u6d4b<\/A> \u592a\u9633\u9ed1\u5b50<\/A>","jumpurl":"","iframeurl":"","style":"","template":"a:3:{s:4:\"head\";s:0:\"\";s:4:\"foot\";s:0:\"\";s:8:\"bencandy\";s:0:\"\";}","target":"0","ip":"120.229.33.54","lastfid":"0","money":"0","buyuser":"","passwd":"","allowdown":"","allowview":"","editer":"","edittime":"0","begintime":"0","endtime":"0","description":"\u65f6\u95f4\u5e8f\u5217\u6df1\u5ea6\u5b66\u4e60\uff1a\u72b6\u6001 LSTM \u6a21\u578b\u9884\u6d4b\u592a\u9633\u9ed1\u5b50","lastview":"1713575022","digg_num":"0","digg_time":"0","forbidcomment":"0","ifvote":"0","heart":"","htmlname":"","city_id":"0"},"page":"1"}
- \u5c06\u72b6\u6001 LSTM \u6a21\u578b\u5e94\u7528\u5230\u8457\u540d\u7684\u592a\u9633\u9ed1\u5b50<\/strong>\u6570\u636e\u96c6\u4e0a\u3002<\/li> \n
- 2 \u6570\u636e<\/a><\/li> \n
- \u5546\u4e1a\u5e94\u7528<\/a><\/li> \n
- \u6559\u7a0b\u6982\u89c8<\/a><\/li> \n