├── README.md ├── RF-regression.ipynb └── BP.ipynb /README.md: -------------------------------------------------------------------------------- 1 | # stock-price-prediction-algorithms 2 | 使用随机森林、bp神经网络、LSTM神经网络、GRU对股票收盘价进行回归预测。Random forest, BP neural network, LSTM neural network and GRU are used to predict the closing price. 3 | 提供了两个版本,四个算法分别的ipynb文件,以及四个算法放置在一起的对比文件。 4 | -------------------------------------------------------------------------------- /RF-regression.ipynb: -------------------------------------------------------------------------------- 1 | { 2 | "cells": [ 3 | { 4 | "cell_type": "code", 5 | "execution_count": 33, 6 | "metadata": {}, 7 | "outputs": [], 8 | "source": [ 9 | "import numpy as np\n", 10 | "import matplotlib.pyplot as plt\n", 11 | "import pandas as pd\n", 12 | "from sklearn import preprocessing\n", 13 | "from sklearn.metrics import mean_squared_error\n", 14 | "from math import sqrt\n", 15 | "#倒入一些必要的库" 16 | ] 17 | }, 18 | { 19 | "cell_type": "code", 20 | "execution_count": 34, 21 | "metadata": {}, 22 | "outputs": [ 23 | { 24 | "data": { 25 | "text/html": [ 26 | "
\n", 27 | "\n", 40 | "\n", 41 | " \n", 42 | " \n", 43 | " \n", 44 | " \n", 45 | " \n", 46 | " \n", 47 | " \n", 48 | " \n", 49 | " \n", 50 | " \n", 51 | " \n", 52 | " \n", 53 | " \n", 54 | " \n", 55 | " \n", 56 | " \n", 57 | " \n", 58 | " \n", 59 | " \n", 60 | " \n", 61 | " \n", 62 | " \n", 63 | " \n", 64 | " \n", 65 | " \n", 66 | " \n", 67 | " \n", 68 | " \n", 69 | " \n", 70 | " \n", 71 | " \n", 72 | " \n", 73 | " \n", 74 | " \n", 75 | " \n", 76 | " \n", 77 | " \n", 78 | " \n", 79 | " \n", 80 | " \n", 81 | " \n", 82 | " \n", 83 | " \n", 84 | " \n", 85 | " \n", 86 | " \n", 87 | "
openlowclosehigh
262727.10427.10427.10487.1055
262737.10477.10387.10387.1047
262747.10397.10397.10467.1047
262757.10487.10437.10457.1049
262767.10497.10497.10577.1057
\n", 88 | "
" 89 | ], 90 | "text/plain": [ 91 | " open low close high\n", 92 | "26272 7.1042 7.1042 7.1048 7.1055\n", 93 | "26273 7.1047 7.1038 7.1038 7.1047\n", 94 | "26274 7.1039 7.1039 7.1046 7.1047\n", 95 | "26275 7.1048 7.1043 7.1045 7.1049\n", 96 | "26276 7.1049 7.1049 7.1057 7.1057" 97 | ] 98 | }, 99 | "execution_count": 34, 100 | "metadata": {}, 101 | "output_type": "execute_result" 102 | } 103 | ], 104 | "source": [ 105 | "#读取数据\n", 106 | "df1=pd.read_csv('data.csv') \n", 107 | "df1=df1.iloc[:,2:]\n", 108 | "df1.tail()" 109 | ] 110 | }, 111 | { 112 | "cell_type": "code", 113 | "execution_count": 35, 114 | "metadata": {}, 115 | "outputs": [ 116 | { 117 | "name": "stdout", 118 | "output_type": "stream", 119 | "text": [ 120 | "(26277, 3)\n", 121 | "(26277,)\n" 122 | ] 123 | } 124 | ], 125 | "source": [ 126 | "#进行数据归一化\n", 127 | "from sklearn import preprocessing\n", 128 | "min_max_scaler = preprocessing.MinMaxScaler()\n", 129 | "df0=min_max_scaler.fit_transform(df1)\n", 130 | "df = pd.DataFrame(df0, columns=df1.columns)\n", 131 | "X=df.iloc[:,:-1]\n", 132 | "y=df.iloc[:,-1]\n", 133 | "print(X.shape)\n", 134 | "print(y.shape)" 135 | ] 136 | }, 137 | { 138 | "cell_type": "code", 139 | "execution_count": 36, 140 | "metadata": {}, 141 | "outputs": [ 142 | { 143 | "name": "stdout", 144 | "output_type": "stream", 145 | "text": [ 146 | "70950\n", 147 | "7881\n", 148 | "23650\n", 149 | "2627\n" 150 | ] 151 | } 152 | ], 153 | "source": [ 154 | "#构造训练集测试集 \n", 155 | "y=pd.DataFrame(y.values,columns=['target'])\n", 156 | "x=X\n", 157 | "input_size=len(x.iloc[1,:])\n", 158 | "cut=len(y)//10#取最后cut=10天为测试集\n", 159 | "X_train, X_test=x.iloc[:-cut],x.iloc[-cut:]#列表的切片操作\n", 160 | "y_train, y_test=y.iloc[:-cut],y.iloc[-cut:]\n", 161 | "X_train,X_test,y_train,y_test=X_train.values,X_test.values,y_train.values,y_test.values\n", 162 | "x.iloc[:-cut]\n", 163 | "print(X_train.size)#通过输出训练集测试集的大小来判断数据格式正确。\n", 164 | "print(X_test.size)\n", 165 | "print(y_train.size)\n", 166 | "print(y_test.size)" 167 | ] 168 | }, 169 | { 170 | "cell_type": "code", 171 | "execution_count": 37, 172 | "metadata": {}, 173 | "outputs": [ 174 | { 175 | "name": "stderr", 176 | "output_type": "stream", 177 | "text": [ 178 | "E:\\anoconda\\lib\\site-packages\\sklearn\\ensemble\\forest.py:245: FutureWarning: The default value of n_estimators will change from 10 in version 0.20 to 100 in 0.22.\n", 179 | " \"10 in version 0.20 to 100 in 0.22.\", FutureWarning)\n", 180 | "E:\\anoconda\\lib\\site-packages\\ipykernel_launcher.py:4: DataConversionWarning: A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples,), for example using ravel().\n", 181 | " after removing the cwd from sys.path.\n" 182 | ] 183 | } 184 | ], 185 | "source": [ 186 | "# 建立随机森林模型 预测\n", 187 | "from sklearn.ensemble import RandomForestRegressor\n", 188 | "rf=RandomForestRegressor()\n", 189 | "model = rf.fit(X_train, y_train)" 190 | ] 191 | }, 192 | { 193 | "cell_type": "code", 194 | "execution_count": 38, 195 | "metadata": {}, 196 | "outputs": [ 197 | { 198 | "data": { 199 | "text/plain": [ 200 | "Text(0.5, 1.0, 'Train Data')" 201 | ] 202 | }, 203 | "execution_count": 38, 204 | "metadata": {}, 205 | "output_type": "execute_result" 206 | }, 207 | { 208 | "data": { 209 | "image/png": "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\n", 210 | "text/plain": [ 211 | "
" 212 | ] 213 | }, 214 | "metadata": { 215 | "needs_background": "light" 216 | }, 217 | "output_type": "display_data" 218 | } 219 | ], 220 | "source": [ 221 | "#在训练集上的拟合结果\n", 222 | "y_train_predict=model.predict(X_train)\n", 223 | "#展示在训练集上的表现 \n", 224 | "draw=pd.concat([pd.DataFrame(y_train),pd.DataFrame(y_train_predict)],axis=1)\n", 225 | "draw.iloc[100:150,0].plot(figsize=(12,6))\n", 226 | "draw.iloc[100:150,1].plot(figsize=(12,6))\n", 227 | "plt.legend(('real', 'predict'),fontsize='15')\n", 228 | "plt.title(\"Train Data\",fontsize='30') #添加标题" 229 | ] 230 | }, 231 | { 232 | "cell_type": "code", 233 | "execution_count": 39, 234 | "metadata": {}, 235 | "outputs": [ 236 | { 237 | "data": { 238 | "text/plain": [ 239 | "Text(0.5, 1.0, 'Test Data')" 240 | ] 241 | }, 242 | "execution_count": 39, 243 | "metadata": {}, 244 | "output_type": "execute_result" 245 | }, 246 | { 247 | "data": { 248 | "image/png": "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\n", 249 | "text/plain": [ 250 | "
" 251 | ] 252 | }, 253 | "metadata": { 254 | "needs_background": "light" 255 | }, 256 | "output_type": "display_data" 257 | } 258 | ], 259 | "source": [ 260 | "#在测试集上的预测\n", 261 | "y_test_predict=model.predict(X_test)\n", 262 | "#展示在测试集上的表现 \n", 263 | "draw=pd.concat([pd.DataFrame(y_test),pd.DataFrame(y_test_predict)],axis=1);\n", 264 | "draw.iloc[200:250,0].plot(figsize=(12,6))\n", 265 | "draw.iloc[200:250,1].plot(figsize=(12,6))\n", 266 | "plt.legend(('real', 'predict'),loc='upper right',fontsize='15')\n", 267 | "plt.title(\"Test Data\",fontsize='30') #添加标题" 268 | ] 269 | }, 270 | { 271 | "cell_type": "code", 272 | "execution_count": 40, 273 | "metadata": {}, 274 | "outputs": [ 275 | { 276 | "name": "stdout", 277 | "output_type": "stream", 278 | "text": [ 279 | "训练集上的MAE/MSE/MAPE\n", 280 | "0.0003261015220079877\n", 281 | "3.761974378011874e-07\n", 282 | "0.11039980586133406\n", 283 | "测试集上的MAE/MSE/MAPE\n", 284 | "0.0006801879768926986\n", 285 | "1.1597457127580965e-06\n", 286 | "0.09498073169728351\n", 287 | "预测涨跌正确: 0.8084539223153084\n", 288 | "训练时间(秒): 5.25\n" 289 | ] 290 | } 291 | ], 292 | "source": [ 293 | "#输出结果\n", 294 | "from sklearn.metrics import mean_absolute_error\n", 295 | "from sklearn.metrics import mean_squared_error\n", 296 | "import math\n", 297 | "def mape(y_true, y_pred):\n", 298 | " return np.mean(np.abs((y_pred - y_true) / y_true)) * 100\n", 299 | "print('训练集上的MAE/MSE/MAPE')\n", 300 | "print(mean_absolute_error(y_train_predict, y_train))\n", 301 | "print(mean_squared_error(y_train_predict, y_train) )\n", 302 | "print(mape(y_train_predict, y_train[:,0]) )\n", 303 | "print('测试集上的MAE/MSE/MAPE')\n", 304 | "print(mean_absolute_error(y_test_predict, y_test))\n", 305 | "print(mean_squared_error(y_test_predict, y_test) )\n", 306 | "print(mape(y_test_predict, y_test[:,0]) )\n", 307 | "y_var_test=y_test[1:]-y_test[:len(y_test)-1]\n", 308 | "y_var_predict=y_test_predict[1:]-y_test_predict[:len(y_test_predict)-1]\n", 309 | "txt=np.zeros(len(y_var_test))\n", 310 | "for i in range(len(y_var_test-1)):\n", 311 | " txt[i]=np.sign(y_var_test[i])==np.sign(y_var_predict[i])\n", 312 | "result=sum(txt)/len(txt)\n", 313 | "print('预测涨跌正确:',result)\n", 314 | "print('训练时间(秒):',5.25)" 315 | ] 316 | }, 317 | { 318 | "cell_type": "code", 319 | "execution_count": null, 320 | "metadata": {}, 321 | "outputs": [], 322 | "source": [] 323 | }, 324 | { 325 | "cell_type": "code", 326 | "execution_count": null, 327 | "metadata": {}, 328 | "outputs": [], 329 | "source": [] 330 | } 331 | ], 332 | "metadata": { 333 | "anaconda-cloud": {}, 334 | "kernelspec": { 335 | "display_name": "Python 3", 336 | "language": "python", 337 | "name": "python3" 338 | }, 339 | "language_info": { 340 | "codemirror_mode": { 341 | "name": "ipython", 342 | "version": 3 343 | }, 344 | "file_extension": ".py", 345 | "mimetype": "text/x-python", 346 | "name": "python", 347 | "nbconvert_exporter": "python", 348 | "pygments_lexer": "ipython3", 349 | "version": "3.7.3" 350 | } 351 | }, 352 | "nbformat": 4, 353 | "nbformat_minor": 1 354 | } 355 | -------------------------------------------------------------------------------- /BP.ipynb: -------------------------------------------------------------------------------- 1 | { 2 | "cells": [ 3 | { 4 | "cell_type": "code", 5 | "execution_count": 25, 6 | "metadata": {}, 7 | "outputs": [], 8 | "source": [ 9 | "import numpy as np\n", 10 | "import pandas as pd\n", 11 | "import matplotlib.pyplot as plt\n", 12 | "#导入必要的库" 13 | ] 14 | }, 15 | { 16 | "cell_type": "code", 17 | "execution_count": 26, 18 | "metadata": {}, 19 | "outputs": [ 20 | { 21 | "data": { 22 | "text/html": [ 23 | "
\n", 24 | "\n", 37 | "\n", 38 | " \n", 39 | " \n", 40 | " \n", 41 | " \n", 42 | " \n", 43 | " \n", 44 | " \n", 45 | " \n", 46 | " \n", 47 | " \n", 48 | " \n", 49 | " \n", 50 | " \n", 51 | " \n", 52 | " \n", 53 | " \n", 54 | " \n", 55 | " \n", 56 | " \n", 57 | " \n", 58 | " \n", 59 | " \n", 60 | " \n", 61 | " \n", 62 | " \n", 63 | " \n", 64 | " \n", 65 | " \n", 66 | " \n", 67 | " \n", 68 | " \n", 69 | " \n", 70 | " \n", 71 | " \n", 72 | " \n", 73 | " \n", 74 | " \n", 75 | " \n", 76 | " \n", 77 | " \n", 78 | " \n", 79 | " \n", 80 | " \n", 81 | " \n", 82 | " \n", 83 | " \n", 84 | "
openlowclosehigh
262727.10427.10427.10487.1055
262737.10477.10387.10387.1047
262747.10397.10397.10467.1047
262757.10487.10437.10457.1049
262767.10497.10497.10577.1057
\n", 85 | "
" 86 | ], 87 | "text/plain": [ 88 | " open low close high\n", 89 | "26272 7.1042 7.1042 7.1048 7.1055\n", 90 | "26273 7.1047 7.1038 7.1038 7.1047\n", 91 | "26274 7.1039 7.1039 7.1046 7.1047\n", 92 | "26275 7.1048 7.1043 7.1045 7.1049\n", 93 | "26276 7.1049 7.1049 7.1057 7.1057" 94 | ] 95 | }, 96 | "execution_count": 26, 97 | "metadata": {}, 98 | "output_type": "execute_result" 99 | } 100 | ], 101 | "source": [ 102 | "#读取数据\n", 103 | "df1=pd.read_csv('data.csv') \n", 104 | "df1=df1.iloc[:,2:]\n", 105 | "df1.tail()" 106 | ] 107 | }, 108 | { 109 | "cell_type": "code", 110 | "execution_count": 27, 111 | "metadata": {}, 112 | "outputs": [ 113 | { 114 | "name": "stdout", 115 | "output_type": "stream", 116 | "text": [ 117 | "(26277, 3)\n", 118 | "(26277,)\n" 119 | ] 120 | } 121 | ], 122 | "source": [ 123 | "#进行数据归一化\n", 124 | "from sklearn import preprocessing\n", 125 | "min_max_scaler = preprocessing.MinMaxScaler()\n", 126 | "df0=min_max_scaler.fit_transform(df1)\n", 127 | "df = pd.DataFrame(df0, columns=df1.columns)\n", 128 | "X=df.iloc[:,:-1]\n", 129 | "y=df.iloc[:,-1]\n", 130 | "print(X.shape)\n", 131 | "print(y.shape)" 132 | ] 133 | }, 134 | { 135 | "cell_type": "code", 136 | "execution_count": 28, 137 | "metadata": {}, 138 | "outputs": [ 139 | { 140 | "name": "stdout", 141 | "output_type": "stream", 142 | "text": [ 143 | "70950\n", 144 | "7881\n", 145 | "23650\n", 146 | "2627\n" 147 | ] 148 | } 149 | ], 150 | "source": [ 151 | "#构造训练集测试集 \n", 152 | "y=pd.DataFrame(y.values,columns=['target'])\n", 153 | "x=X\n", 154 | "input_size=len(x.iloc[1,:])\n", 155 | "cut=len(y)//10#取最后cut=10天为测试集\n", 156 | "X_train, X_test=x.iloc[:-cut],x.iloc[-cut:]#列表的切片操作\n", 157 | "y_train, y_test=y.iloc[:-cut],y.iloc[-cut:]\n", 158 | "X_train,X_test,y_train,y_test=X_train.values,X_test.values,y_train.values,y_test.values\n", 159 | "x.iloc[:-cut]\n", 160 | "print(X_train.size)#通过输出训练集测试集的大小来判断数据格式正确。\n", 161 | "print(X_test.size)\n", 162 | "print(y_train.size)\n", 163 | "print(y_test.size)" 164 | ] 165 | }, 166 | { 167 | "cell_type": "code", 168 | "execution_count": 29, 169 | "metadata": {}, 170 | "outputs": [ 171 | { 172 | "name": "stderr", 173 | "output_type": "stream", 174 | "text": [ 175 | "E:\\anoconda\\lib\\site-packages\\ipykernel_launcher.py:6: UserWarning: Update your `Dense` call to the Keras 2 API: `Dense(16, input_dim=3, kernel_initializer=\"uniform\")`\n", 176 | " \n", 177 | "E:\\anoconda\\lib\\site-packages\\ipykernel_launcher.py:8: UserWarning: Update your `Dense` call to the Keras 2 API: `Dense(4, kernel_initializer=\"uniform\")`\n", 178 | " \n", 179 | "E:\\anoconda\\lib\\site-packages\\ipykernel_launcher.py:12: UserWarning: The `nb_epoch` argument in `fit` has been renamed `epochs`.\n", 180 | " if sys.path[0] == '':\n" 181 | ] 182 | }, 183 | { 184 | "name": "stdout", 185 | "output_type": "stream", 186 | "text": [ 187 | "Epoch 1/50\n", 188 | "23650/23650 [==============================] - 0s 15us/step - loss: 0.2026\n", 189 | "Epoch 2/50\n", 190 | "23650/23650 [==============================] - 0s 6us/step - loss: 0.0248\n", 191 | "Epoch 3/50\n", 192 | "23650/23650 [==============================] - 0s 7us/step - loss: 0.0127\n", 193 | "Epoch 4/50\n", 194 | "23650/23650 [==============================] - 0s 8us/step - loss: 0.0085\n", 195 | "Epoch 5/50\n", 196 | "23650/23650 [==============================] - 0s 7us/step - loss: 0.0048\n", 197 | "Epoch 6/50\n", 198 | "23650/23650 [==============================] - 0s 6us/step - loss: 0.0023\n", 199 | "Epoch 7/50\n", 200 | "23650/23650 [==============================] - 0s 5us/step - loss: 8.6696e-04\n", 201 | "Epoch 8/50\n", 202 | "23650/23650 [==============================] - 0s 6us/step - loss: 2.7231e-04\n", 203 | "Epoch 9/50\n", 204 | "23650/23650 [==============================] - 0s 5us/step - loss: 8.1726e-05\n", 205 | "Epoch 10/50\n", 206 | "23650/23650 [==============================] - 0s 7us/step - loss: 3.1396e-05\n", 207 | "Epoch 11/50\n", 208 | "23650/23650 [==============================] - 0s 6us/step - loss: 2.0191e-05\n", 209 | "Epoch 12/50\n", 210 | "23650/23650 [==============================] - 0s 6us/step - loss: 1.7813e-05\n", 211 | "Epoch 13/50\n", 212 | "23650/23650 [==============================] - 0s 6us/step - loss: 1.7124e-05\n", 213 | "Epoch 14/50\n", 214 | "23650/23650 [==============================] - 0s 8us/step - loss: 1.6705e-05\n", 215 | "Epoch 15/50\n", 216 | "23650/23650 [==============================] - 0s 6us/step - loss: 1.6323e-05\n", 217 | "Epoch 16/50\n", 218 | "23650/23650 [==============================] - 0s 6us/step - loss: 1.5953e-05\n", 219 | "Epoch 17/50\n", 220 | "23650/23650 [==============================] - 0s 6us/step - loss: 1.5589e-05\n", 221 | "Epoch 18/50\n", 222 | "23650/23650 [==============================] - 0s 6us/step - loss: 1.5223e-05\n", 223 | "Epoch 19/50\n", 224 | "23650/23650 [==============================] - 0s 6us/step - loss: 1.4864e-05\n", 225 | "Epoch 20/50\n", 226 | "23650/23650 [==============================] - 0s 9us/step - loss: 1.4503e-05\n", 227 | "Epoch 21/50\n", 228 | "23650/23650 [==============================] - 0s 9us/step - loss: 1.4152e-05\n", 229 | "Epoch 22/50\n", 230 | "23650/23650 [==============================] - 0s 6us/step - loss: 1.3811e-05\n", 231 | "Epoch 23/50\n", 232 | "23650/23650 [==============================] - 0s 5us/step - loss: 1.3477e-05\n", 233 | "Epoch 24/50\n", 234 | "23650/23650 [==============================] - 0s 4us/step - loss: 1.3133e-05\n", 235 | "Epoch 25/50\n", 236 | "23650/23650 [==============================] - 0s 4us/step - loss: 1.2810e-05\n", 237 | "Epoch 26/50\n", 238 | "23650/23650 [==============================] - 0s 5us/step - loss: 1.2497e-05\n", 239 | "Epoch 27/50\n", 240 | "23650/23650 [==============================] - 0s 6us/step - loss: 1.2193e-05\n", 241 | "Epoch 28/50\n", 242 | "23650/23650 [==============================] - 0s 6us/step - loss: 1.1889e-05\n", 243 | "Epoch 29/50\n", 244 | "23650/23650 [==============================] - 0s 5us/step - loss: 1.1593e-05\n", 245 | "Epoch 30/50\n", 246 | "23650/23650 [==============================] - 0s 4us/step - loss: 1.1298e-05\n", 247 | "Epoch 31/50\n", 248 | "23650/23650 [==============================] - ETA: 0s - loss: 1.0749e-0 - 0s 6us/step - loss: 1.1015e-05\n", 249 | "Epoch 32/50\n", 250 | "23650/23650 [==============================] - 0s 5us/step - loss: 1.0739e-05\n", 251 | "Epoch 33/50\n", 252 | "23650/23650 [==============================] - 0s 5us/step - loss: 1.0469e-05\n", 253 | "Epoch 34/50\n", 254 | "23650/23650 [==============================] - 0s 5us/step - loss: 1.0202e-05\n", 255 | "Epoch 35/50\n", 256 | "23650/23650 [==============================] - 0s 4us/step - loss: 9.9408e-06\n", 257 | "Epoch 36/50\n", 258 | "23650/23650 [==============================] - 0s 5us/step - loss: 9.6897e-06\n", 259 | "Epoch 37/50\n", 260 | "23650/23650 [==============================] - 0s 5us/step - loss: 9.4334e-06\n", 261 | "Epoch 38/50\n", 262 | "23650/23650 [==============================] - 0s 5us/step - loss: 9.1941e-06\n", 263 | "Epoch 39/50\n", 264 | "23650/23650 [==============================] - 0s 5us/step - loss: 8.9507e-06\n", 265 | "Epoch 40/50\n", 266 | "23650/23650 [==============================] - 0s 5us/step - loss: 8.7264e-06\n", 267 | "Epoch 41/50\n", 268 | "23650/23650 [==============================] - 0s 5us/step - loss: 8.5088e-06\n", 269 | "Epoch 42/50\n", 270 | "23650/23650 [==============================] - 0s 5us/step - loss: 8.2959e-06\n", 271 | "Epoch 43/50\n", 272 | "23650/23650 [==============================] - 0s 5us/step - loss: 8.0945e-06\n", 273 | "Epoch 44/50\n", 274 | "23650/23650 [==============================] - 0s 5us/step - loss: 7.8819e-06\n", 275 | "Epoch 45/50\n", 276 | "23650/23650 [==============================] - 0s 5us/step - loss: 7.6909e-06A: 0s - loss: 7.7156e-0\n", 277 | "Epoch 46/50\n", 278 | "23650/23650 [==============================] - 0s 4us/step - loss: 7.4975e-06\n", 279 | "Epoch 47/50\n", 280 | "23650/23650 [==============================] - 0s 5us/step - loss: 7.3296e-06\n", 281 | "Epoch 48/50\n", 282 | "23650/23650 [==============================] - 0s 6us/step - loss: 7.1390e-06\n", 283 | "Epoch 49/50\n", 284 | "23650/23650 [==============================] - 0s 4us/step - loss: 6.9767e-06\n", 285 | "Epoch 50/50\n", 286 | "23650/23650 [==============================] - 0s 5us/step - loss: 6.8290e-06\n" 287 | ] 288 | }, 289 | { 290 | "data": { 291 | "text/plain": [ 292 | "" 293 | ] 294 | }, 295 | "execution_count": 29, 296 | "metadata": {}, 297 | "output_type": "execute_result" 298 | } 299 | ], 300 | "source": [ 301 | "#建立bp模型 训练 \n", 302 | "from keras.models import Sequential\n", 303 | "from keras.layers.core import Dense, Activation\n", 304 | "from keras.optimizers import Adam\n", 305 | "model = Sequential() #层次模型\n", 306 | "model.add(Dense(16,input_dim=input_size,init='uniform')) #输入层,Dense表示BP层\n", 307 | "model.add(Activation('relu')) #添加激活函数\n", 308 | "model.add(Dense(4,init='uniform')) #中间层\n", 309 | "model.add(Activation('relu')) #添加激活函数\n", 310 | "model.add(Dense(1)) #输出层\n", 311 | "model.compile(loss='mean_squared_error', optimizer='Adam') #编译模型\n", 312 | "model.fit(X_train, y_train, nb_epoch = 50, batch_size = 256) #训练模型nb_epoch=50次" 313 | ] 314 | }, 315 | { 316 | "cell_type": "code", 317 | "execution_count": 30, 318 | "metadata": {}, 319 | "outputs": [ 320 | { 321 | "name": "stdout", 322 | "output_type": "stream", 323 | "text": [ 324 | "Model: \"sequential_2\"\n", 325 | "_________________________________________________________________\n", 326 | "Layer (type) Output Shape Param # \n", 327 | "=================================================================\n", 328 | "dense_4 (Dense) (None, 16) 64 \n", 329 | "_________________________________________________________________\n", 330 | "activation_3 (Activation) (None, 16) 0 \n", 331 | "_________________________________________________________________\n", 332 | "dense_5 (Dense) (None, 4) 68 \n", 333 | "_________________________________________________________________\n", 334 | "activation_4 (Activation) (None, 4) 0 \n", 335 | "_________________________________________________________________\n", 336 | "dense_6 (Dense) (None, 1) 5 \n", 337 | "=================================================================\n", 338 | "Total params: 137\n", 339 | "Trainable params: 137\n", 340 | "Non-trainable params: 0\n", 341 | "_________________________________________________________________\n" 342 | ] 343 | } 344 | ], 345 | "source": [ 346 | "model.summary()#模型描述" 347 | ] 348 | }, 349 | { 350 | "cell_type": "code", 351 | "execution_count": 31, 352 | "metadata": {}, 353 | "outputs": [], 354 | "source": [ 355 | "#在训练集上的拟合结果\n", 356 | "y_train_predict=model.predict(X_train)\n", 357 | "y_train_predict=y_train_predict[:,0]\n", 358 | "y_train=y_train" 359 | ] 360 | }, 361 | { 362 | "cell_type": "code", 363 | "execution_count": 32, 364 | "metadata": {}, 365 | "outputs": [ 366 | { 367 | "data": { 368 | "text/plain": [ 369 | "Text(0.5, 1.0, 'Train Data')" 370 | ] 371 | }, 372 | "execution_count": 32, 373 | "metadata": {}, 374 | "output_type": "execute_result" 375 | }, 376 | { 377 | "data": { 378 | "image/png": "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\n", 379 | "text/plain": [ 380 | "
" 381 | ] 382 | }, 383 | "metadata": { 384 | "needs_background": "light" 385 | }, 386 | "output_type": "display_data" 387 | } 388 | ], 389 | "source": [ 390 | "draw=pd.concat([pd.DataFrame(y_train),pd.DataFrame(y_train_predict)],axis=1)\n", 391 | "draw.iloc[100:150,0].plot(figsize=(12,6))\n", 392 | "draw.iloc[100:150,1].plot(figsize=(12,6))\n", 393 | "plt.legend(('real', 'predict'),fontsize='15')\n", 394 | "plt.title(\"Train Data\",fontsize='30') #添加标题\n", 395 | "#展示在训练集上的表现 " 396 | ] 397 | }, 398 | { 399 | "cell_type": "code", 400 | "execution_count": 33, 401 | "metadata": {}, 402 | "outputs": [], 403 | "source": [ 404 | "#在测试集上的预测\n", 405 | "y_test_predict=model.predict(X_test)\n", 406 | "y_test_predict=y_test_predict[:,0]" 407 | ] 408 | }, 409 | { 410 | "cell_type": "code", 411 | "execution_count": 34, 412 | "metadata": {}, 413 | "outputs": [ 414 | { 415 | "data": { 416 | "text/plain": [ 417 | "Text(0.5, 1.0, 'Test Data')" 418 | ] 419 | }, 420 | "execution_count": 34, 421 | "metadata": {}, 422 | "output_type": "execute_result" 423 | }, 424 | { 425 | "data": { 426 | "image/png": "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\n", 427 | "text/plain": [ 428 | "
" 429 | ] 430 | }, 431 | "metadata": { 432 | "needs_background": "light" 433 | }, 434 | "output_type": "display_data" 435 | } 436 | ], 437 | "source": [ 438 | "draw=pd.concat([pd.DataFrame(y_test),pd.DataFrame(y_test_predict)],axis=1);\n", 439 | "draw.iloc[200:250,0].plot(figsize=(12,6))\n", 440 | "draw.iloc[200:250,1].plot(figsize=(12,6))\n", 441 | "plt.legend(('real', 'predict'),loc='upper right',fontsize='15')\n", 442 | "plt.title(\"Test Data\",fontsize='30') #添加标题\n", 443 | "#展示在测试集上的表现 " 444 | ] 445 | }, 446 | { 447 | "cell_type": "code", 448 | "execution_count": 35, 449 | "metadata": {}, 450 | "outputs": [ 451 | { 452 | "name": "stdout", 453 | "output_type": "stream", 454 | "text": [ 455 | "训练集上的MAE/MSE/MAPE\n", 456 | "0.0015122720594528372\n", 457 | "6.715321538274574e-06\n", 458 | "0.595540765012317\n", 459 | "测试集上的MAE/MSE/MAPE\n", 460 | "0.001278273879630919\n", 461 | "2.6149447835062204e-06\n", 462 | "0.17901381364811333\n", 463 | "预测涨跌正确: 0.8175932977913176\n", 464 | "训练时间(秒): 15.25\n" 465 | ] 466 | } 467 | ], 468 | "source": [ 469 | "#输出结果\n", 470 | "from sklearn.metrics import mean_absolute_error\n", 471 | "from sklearn.metrics import mean_squared_error\n", 472 | "import math\n", 473 | "def mape(y_true, y_pred):\n", 474 | " return np.mean(np.abs((y_pred - y_true) / y_true)) * 100\n", 475 | "print('训练集上的MAE/MSE/MAPE')\n", 476 | "print(mean_absolute_error(y_train_predict, y_train))\n", 477 | "print(mean_squared_error(y_train_predict, y_train) )\n", 478 | "print(mape(y_train_predict, y_train[:,0]) )\n", 479 | "print('测试集上的MAE/MSE/MAPE')\n", 480 | "print(mean_absolute_error(y_test_predict, y_test))\n", 481 | "print(mean_squared_error(y_test_predict, y_test) )\n", 482 | "print(mape(y_test_predict, y_test[:,0]) )\n", 483 | "y_var_test=y_test[1:]-y_test[:len(y_test)-1]\n", 484 | "y_var_predict=y_test_predict[1:]-y_test_predict[:len(y_test_predict)-1]\n", 485 | "txt=np.zeros(len(y_var_test))\n", 486 | "for i in range(len(y_var_test-1)):\n", 487 | " txt[i]=np.sign(y_var_test[i])==np.sign(y_var_predict[i])\n", 488 | "result=sum(txt)/len(txt)\n", 489 | "print('预测涨跌正确:',result)\n", 490 | "print('训练时间(秒):',15.25)" 491 | ] 492 | }, 493 | { 494 | "cell_type": "code", 495 | "execution_count": null, 496 | "metadata": {}, 497 | "outputs": [], 498 | "source": [] 499 | }, 500 | { 501 | "cell_type": "code", 502 | "execution_count": null, 503 | "metadata": {}, 504 | "outputs": [], 505 | "source": [] 506 | }, 507 | { 508 | "cell_type": "code", 509 | "execution_count": null, 510 | "metadata": {}, 511 | "outputs": [], 512 | "source": [] 513 | } 514 | ], 515 | "metadata": { 516 | "kernelspec": { 517 | "display_name": "Python 3", 518 | "language": "python", 519 | "name": "python3" 520 | }, 521 | "language_info": { 522 | "codemirror_mode": { 523 | "name": "ipython", 524 | "version": 3 525 | }, 526 | "file_extension": ".py", 527 | "mimetype": "text/x-python", 528 | "name": "python", 529 | "nbconvert_exporter": "python", 530 | "pygments_lexer": "ipython3", 531 | "version": "3.7.3" 532 | } 533 | }, 534 | "nbformat": 4, 535 | "nbformat_minor": 2 536 | } 537 | --------------------------------------------------------------------------------