\ud000\ud2b8 \ub9e4\ub9e4\uc758 \uc2dc\ub300\uac00 \ub3c4\ub798\ud558\uba74\uc11c, \ub9ce\uc740 \ud22c\uc790\uc790\ub4e4\uc740 \uc54c\uace0\ub9ac\uc998 \ud2b8\ub808\uc774\ub529\uc744 \ud1b5\ud574 \uc2dc\uc7a5\uc5d0\uc11c\uc758 \uacbd\uc7c1\ub825\uc744 \ub192\uc774\uace0 \uc788\uc2b5\ub2c8\ub2e4. \uc774 \uacfc\uc815\uc5d0\uc11c \uba38\uc2e0\ub7ec\ub2dd\uacfc \ub525\ub7ec\ub2dd \uae30\uc220\uc740 \ub9e4\uc6b0 \uc911\uc694\ud55c \uc5ed\ud560\uc744 \ud558\uace0 \uc788\uc73c\uba70, \ud2b9\ud788 \uc9d1\ub77c\uc778(Zipline)\uacfc \uac19\uc740 \ud504\ub808\uc784\uc6cc\ud06c\uac00 \uc774\ub4e4\uc758 \ud65c\uc6a9\uc744 \ubcf4\ub2e4 \uc27d\uac8c \ub9cc\ub4e4\uc5b4 \uc8fc\uace0 \uc788\uc2b5\ub2c8\ub2e4. \uc774\ubc88 \uac15\uc88c\uc5d0\uc11c\ub294 \uba38\uc2e0\ub7ec\ub2dd \ubc0f \ub525\ub7ec\ub2dd \uc54c\uace0\ub9ac\uc998 \ud2b8\ub808\uc774\ub529\uc758 \uae30\ucd08\ubd80\ud130 \uc2dc\uc791\ud558\uc5ec, \uc9d1\ub77c\uc778\uc744 \uc0ac\uc6a9\ud55c \ubc31\ud14c\uc2a4\ud2b8 \uae30\ubc95\uae4c\uc9c0 \uc0c1\uc138\ud788 \uc124\uba85\ub4dc\ub9ac\uaca0\uc2b5\ub2c8\ub2e4.<\/p>\n
\ud000\ud2b8 \ud2b8\ub808\uc774\ub529(Quantitative Trading)\uc774\ub780 \uc218\ud559\uc801 \ubaa8\ub378\uacfc \ud1b5\uacc4\uc801 \uae30\ubc95\uc744 \uc0ac\uc6a9\ud558\uc5ec \uae08\uc735 \uc2dc\uc7a5\uc5d0\uc11c \uac70\ub798\ub97c \uc218\ud589\ud558\ub294 \ubc29\uc2dd\uc785\ub2c8\ub2e4. \uc774 \uacfc\uc815\uc5d0\uc11c\ub294 \ub300\uaddc\ubaa8 \ub370\uc774\ud130 \ubd84\uc11d\uacfc \uc54c\uace0\ub9ac\uc998 \uc791\uc131\uc744 \ud1b5\ud574 \ucd5c\uc801\uc758 \uac70\ub798 \uc804\ub7b5\uc744 \uc218\ub9bd\ud558\uac8c \ub429\ub2c8\ub2e4.<\/p>\n
\uc804\ud1b5\uc801\uc778 \ud000\ud2b8 \ud2b8\ub808\uc774\ub529 \uae30\ubc95\uc740 \ub300\ubd80\ubd84 \uace0\uc815\ub41c \uaddc\uce59\uc5d0 \uae30\ubc18\ud558\uc5ec \uc791\ub3d9\ud558\uc9c0\ub9cc, \uba38\uc2e0\ub7ec\ub2dd\uc740 \ub370\uc774\ud130\uc5d0\uc11c \ud328\ud134\uc744 \uc790\ub3d9\uc73c\ub85c \ud559\uc2b5\ud558\uace0 \uac1c\uc120\ud560 \uc218 \uc788\uc2b5\ub2c8\ub2e4. \uadf8 \uacb0\uacfc, \uc2dc\uc7a5\uc758 \ubcc0\ud654\ub97c \ub354 \uc798 \ubc18\uc601\ud55c \uc608\uce21 \ubaa8\ub378\uc744 \uad6c\ucd95\ud560 \uc218 \uc788\uac8c \ub429\ub2c8\ub2e4.<\/p>\n
\ub525\ub7ec\ub2dd\uc740 \uba38\uc2e0\ub7ec\ub2dd\uc758 \ud55c \ubd84\uc57c\ub85c, \uc778\uacf5 \uc2e0\uacbd\ub9dd\uc744 \uc774\uc6a9\ud558\uc5ec \ub370\uc774\ud130\uc758 \ubcf5\uc7a1\ud55c \ud328\ud134\uc744 \uc778\uc2dd\ud569\ub2c8\ub2e4. \ud2b9\ud788, \ub300\ub7c9\uc758 \ube44\uc815\ud615 \ub370\uc774\ud130(\uc608: \ub274\uc2a4 \uae30\uc0ac, \uc18c\uc15c \ubbf8\ub514\uc5b4 \ub370\uc774\ud130)\uc5d0\uc11c \uc720\uc6a9\ud55c \uc778\uc0ac\uc774\ud2b8\ub97c \ucd94\ucd9c\ud560 \uc218 \uc788\uc2b5\ub2c8\ub2e4.<\/p>\n
\uc9d1\ub77c\uc778\uc740 \ud30c\uc774\uc36c \uae30\ubc18\uc758 \uc624\ud508 \uc18c\uc2a4 \ubc31\ud14c\uc2a4\ud305 \ub77c\uc774\ube0c\ub7ec\ub9ac\ub85c, \ud000\ud2b8 \uc804\ub7b5\uc758 \uac1c\ubc1c\uacfc \ud14c\uc2a4\ud2b8\uc5d0 \uad11\ubc94\uc704\ud558\uac8c \uc0ac\uc6a9\ub418\uace0 \uc788\uc2b5\ub2c8\ub2e4. \uc0ac\uc6a9\uc790\uac00 \uc815\uc758\ud55c \uc54c\uace0\ub9ac\uc998\uc744 \ubc14\ud0d5\uc73c\ub85c \uc5ed\uc0ac\uc801 \ub370\uc774\ud130\ub97c \uc774\uc6a9\ud558\uc5ec \uc804\ub7b5\uc758 \ud6a8\uc728\uc131\uc744 \ud3c9\uac00\ud560 \uc218 \uc788\uc2b5\ub2c8\ub2e4.<\/p>\n
\uc6b0\uc120, \ud544\uc694\ud55c \ub370\uc774\ud130\ub97c \uc218\uc9d1\ud574\uc57c \ud569\ub2c8\ub2e4. Yahoo Finance, Alpha Vantage, Quandl\uacfc \uac19\uc740 \ud50c\ub7ab\ud3fc\uc5d0\uc11c \uae08\uc735 \uad00\ub828 \ub370\uc774\ud130\ub97c API\ub97c \uc774\uc6a9\ud558\uc5ec \uc218\uc9d1\ud560 \uc218 \uc788\uc2b5\ub2c8\ub2e4. \uc774 \ub370\uc774\ud130\ub294 \ubaa8\ub378 \ud559\uc2b5\uc758 \uae30\ucd08\uac00 \ub429\ub2c8\ub2e4.<\/p>\n
\uc218\uc9d1\ud55c \ub370\uc774\ud130\ub294 \ud56d\uc0c1 \uae68\ub057\ud558\uc9c0 \uc54a\uc73c\uba70, \uc804\ucc98\ub9ac \uacfc\uc815\uc744 \ud1b5\ud574 \uc815\uc81c\ub418\uc5b4\uc57c \ud569\ub2c8\ub2e4. \uacb0\uce21\uac12 \ucc98\ub9ac, \uc815\uaddc\ud654, \ub808\uc774\ube14 \uc778\ucf54\ub529 \ub4f1\uc758 \uacfc\uc815\uc744 \ud1b5\ud574 \uba38\uc2e0\ub7ec\ub2dd \ubaa8\ub378\uc774 \uc774\ud574\ud560 \uc218 \uc788\ub294 \ud615\ud0dc\ub85c \ubcc0\ud658\ud569\ub2c8\ub2e4.<\/p>\n
\ubaa8\ub378\uc758 \uc131\ub2a5\uc744 \ub192\uc774\uae30 \uc704\ud574\uc11c\ub294 \uc720\uc758\ubbf8\ud55c \ud2b9\uc131\uc744 \uc120\ud0dd\ud558\ub294 \uac83\uc774 \uc911\uc694\ud569\ub2c8\ub2e4. \uae08\uc735 \uc2dc\uc7a5\uc5d0\uc11c\ub294 \uc774\ub3d9 \ud3c9\uade0, RSI, MACD \ub4f1\uc758 \uc9c0\ud45c\ub97c \ud2b9\uc131\uc73c\ub85c \uc0ac\uc6a9\ud560 \uc218 \uc788\uc2b5\ub2c8\ub2e4.<\/p>\n
\uba38\uc2e0\ub7ec\ub2dd \ubaa8\ub378\ub85c\ub294 \ud68c\uadc0, \uc758\uc0ac\uacb0\uc815 \ub098\ubb34, \ub79c\ub364 \ud3ec\ub808\uc2a4\ud2b8, XGBoost \ub4f1\uc774 \uc788\uc73c\uba70, \ub525\ub7ec\ub2dd\uc5d0\uc11c\ub294 LSTM, CNN \uac19\uc740 \ubaa8\ub378\uc744 \uc0ac\uc6a9\ud560 \uc218 \uc788\uc2b5\ub2c8\ub2e4. \ucd5c\uc801\uc758 \ubaa8\ub378\uc744 \uc120\uc815\ud558\uace0, \ub370\uc774\ud130\ub97c Training\uacfc Validation set\uc73c\ub85c \ub098\ub204\uc5b4 \ud559\uc2b5\ud569\ub2c8\ub2e4.<\/p>\n
\ubaa8\ub378\uc758 \uc131\uacfc\ub97c \ud3c9\uac00\ud558\uae30 \uc704\ud574 MSE, RMSE, Accuracy, F1 Score \ub4f1 \ub2e4\uc591\ud55c \uc9c0\ud45c\ub97c \uc0ac\uc6a9\ud560 \uc218 \uc788\uc2b5\ub2c8\ub2e4. \uc774 \uacfc\uc815\uc5d0\uc11c \uacfc\uc801\ud569(overfitting) \ubb38\uc81c\ub97c \ubc29\uc9c0\ud558\uae30 \uc704\ud55c \ud06c\ub85c\uc2a4 \ubc38\ub9ac\ub370\uc774\uc158\uc744 \uc801\uc6a9\ud558\ub294 \uac83\uc774 \uc88b\uc2b5\ub2c8\ub2e4.<\/p>\n
\uc9d1\ub77c\uc778\uc744 \uc124\uce58\ud558\uae30 \uc704\ud574\uc11c\ub294 \uc9d1\ub77c\uc778\uc5d0\uc11c\ub294 \uc54c\uace0\ub9ac\uc998\uc744 \ubc31\ud14c\uc2a4\ud2b8 \uacb0\uacfc\ub294 \uc9d1\ub77c\uc778\uc758 \ud559\uc2b5\ud55c \uba38\uc2e0\ub7ec\ub2dd \ubaa8\ub378\uc744 \uc774\uc6a9\ud558\uc5ec \ub9e4\ub9e4 \uc2e0\ud638\ub97c \ubc1c\uc0dd\uc2dc\ud0ac \uc218 \uc788\uc2b5\ub2c8\ub2e4. scikit-learn\uacfc \uac19\uc740 \ub77c\uc774\ube0c\ub7ec\ub9ac\ub97c \ud1b5\ud574 \ubaa8\ub378\uc744 \ud6c8\ub828\uc2dc\ud0a8 \ud6c4, \uc774\ubc88 \uac15\uc88c\uc5d0\uc11c\ub294 \uba38\uc2e0\ub7ec\ub2dd \ubc0f \ub525\ub7ec\ub2dd \uae30\ubc18 \uc54c\uace0\ub9ac\uc998 \ud2b8\ub808\uc774\ub529\uc758 \uae30\ucd08\uc640 \uc9d1\ub77c\uc778\uc744 \ud1b5\ud55c \ubc31\ud14c\uc2a4\ud2b8 \ubc29\ubc95\uc5d0 \ub300\ud574 \uc54c\uc544\ubcf4\uc558\uc2b5\ub2c8\ub2e4. \ud000\ud2b8 \ud2b8\ub808\uc774\ub529\uc740 \uc810\uc810 \ub354 \ubcf5\uc7a1\ud574\uc9c0\uace0 \uc788\uc73c\uba70, \uba38\uc2e0\ub7ec\ub2dd\uacfc \ub525\ub7ec\ub2dd \uae30\uc220\uc744 \uacb0\ud569\ud558\uba74 \ub354 \ub098\uc740 \uc608\uce21\uacfc \uc758\uc0ac\uacb0\uc815\uc744 \ud560 \uc218 \uc788\ub294 \uc7a0\uc7ac\ub825\uc774 \ud07d\ub2c8\ub2e4. \uc55e\uc73c\ub85c\ub294 \ub370\uc774\ud130 \ubd84\uc11d \uae30\ubc95, \ub2e4\uc591\ud55c \ubaa8\ub378 \ud0d0\uc0c9 \ubc0f \uc131\uacfc \ud3c9\uac00 \ubc29\ubc95 \ub4f1\uc744 \uc2ec\ub3c4 \uc788\uac8c \ub2e4\ub8e8\uc5b4 \ub098\uac08 \uacc4\ud68d\uc785\ub2c8\ub2e4.<\/p>\n \ub3c5\uc790 \uc5ec\ub7ec\ubd84\ub4e4\uc774 \uc54c\uace0\ub9ac\uc998 \ud2b8\ub808\uc774\ub529\uc758 \uc138\uacc4\uc5d0 \uc131\uacf5\uc801\uc73c\ub85c \uc9c4\uc785\ud558\uae38 \ubc14\ub77c\uba70, \uafb8\uc900\ud55c \ud559\uc2b5\uacfc \uc2e4\ud5d8\uc744 \ud1b5\ud574 \uc790\uc2e0\uc758 \uc804\ub7b5\uc744 \uac1c\ubc1c\ud574 \ub098\uac00\uc2dc\uae30\ub97c \ubc14\ub78d\ub2c8\ub2e4.<\/p>\n <\/body><\/p>\n","protected":false},"excerpt":{"rendered":"pip install zipline<\/code> \uba85\ub839\uc5b4\ub97c \uc0ac\uc6a9\ud569\ub2c8\ub2e4. \uc8fc\uc758\ud560 \uc810\uc740, Ubuntu\uc640 \uac19\uc740 \ub9ac\ub205\uc2a4 \ud658\uacbd\uc5d0\uc11c \uac00\uc7a5 \uc798 \uc791\ub3d9\ud558\uba70, \uc708\ub3c4\uc6b0 \ud658\uacbd\uc5d0\uc11c\uc758 \uc124\uce58\ub294 \uc81c\uc57d\uc774 \uc788\uc744 \uc218 \uc788\uc2b5\ub2c8\ub2e4.<\/p>\n4.2 \uae30\ubcf8\uc801\uc778 \uc9d1\ub77c\uc778 \uad6c\uc870<\/h3>\n
initialize()<\/code>\uc640 handle_data()<\/code> \ud568\uc218\ub85c \ub098\ub204\uc5b4 \uc791\uc131\ud569\ub2c8\ub2e4. initialize()<\/code>\uc5d0\uc11c\ub294 \ucd08\uae30 \ud30c\ub77c\ubbf8\ud130\uc640 \ubcc0\uc218\ub97c \uc124\uc815\ud558\uace0, handle_data()<\/code>\uc5d0\uc11c\ub294 \ub9e4 \uac70\ub798\uc77c\ub9c8\ub2e4 \uc2e4\ud589\ub418\ub294 \ub85c\uc9c1\uc744 \uc218\ub9bd\ud569\ub2c8\ub2e4.<\/p>\n4.3 \uc608\uc81c \ucf54\ub4dc: \uac04\ub2e8\ud55c \uc774\ub3d9 \ud3c9\uade0 \uad50\ucc28 \uc804\ub7b5<\/h3>\n
\nfrom zipline.api import order, record, symbol\nfrom zipline import run_algorithm\nimport pandas as pd\nfrom datetime import datetime\n\ndef initialize(context):\n context.asset = symbol('AAPL')\n context.short_window = 40\n context.long_window = 100\n\ndef handle_data(context, data):\n # \uc5ed\uc0ac\uc801 \uac00\uaca9 \ub370\uc774\ud130 \uac00\uc838\uc624\uae30\n prices = data.history(context.asset, 'price', context.long_window, '1d')\n \n # \uc774\ub3d9 \ud3c9\uade0 \uacc4\uc0b0\n short_mavg = prices[-context.short_window:].mean()\n long_mavg = prices.mean()\n \n # \ub9e4\uc218 \/ \ub9e4\ub3c4\uc758 \uc870\uac74\n if short_mavg > long_mavg:\n order(context.asset, 1)\n elif short_mavg < long_mavg:\n order(context.asset, -1)\n \n # \uae30\ub85d\n record(AAPL=data.current(context.asset, 'price'))\n\n# \ubc31\ud14c\uc2a4\ud2b8 \uc2e4\ud589\nstart = datetime(2015, 1, 1)\nend = datetime(2017, 1, 1)\nrun_algorithm(start=start, end=end, initialize=initialize, handle_data=handle_data)\n<\/code><\/pre>\n4.4 \uacb0\uacfc \ubd84\uc11d<\/h3>\n
record<\/code>\ub97c \ud1b5\ud574 \uc218\uc9d1\ud560 \uc218 \uc788\uc73c\uba70, \uc2dc\uac01\ud654\ub97c \ud1b5\ud574 \uc131\uacfc\ub97c \ubd84\uc11d\ud560 \uc218 \uc788\uc2b5\ub2c8\ub2e4. \uc774\ub97c \uc704\ud574 matplotlib<\/code>\uc640 \uac19\uc740 \ub77c\uc774\ube0c\ub7ec\ub9ac\ub97c \ud65c\uc6a9\ud558\ub294 \uac83\uc774 \uc88b\uc2b5\ub2c8\ub2e4.<\/p>\n5. \uba38\uc2e0\ub7ec\ub2dd \ubaa8\ub378\uc744 \uc9d1\ub77c\uc778\uc5d0 \ud1b5\ud569\ud558\uae30<\/h2>\n
5.1 \uba38\uc2e0\ub7ec\ub2dd \ubaa8\ub378 \ud6c8\ub828 \ubc0f \uc608\uce21<\/h3>\n
handle_data()<\/code> \ud568\uc218 \ub0b4\uc5d0\uc11c \uc608\uce21 \uacb0\uacfc\ub97c \ud65c\uc6a9\ud558\uc5ec \uc8fc\ubb38 \uacb0\uc815\uc744 \ub0b4\ub9bd\ub2c8\ub2e4.<\/p>\n5.2 \uc608\uc81c \ucf54\ub4dc: \uba38\uc2e0\ub7ec\ub2dd\uacfc \uc9d1\ub77c\uc778 \ud1b5\ud569<\/h3>\n
\nfrom sklearn.ensemble import RandomForestClassifier\nfrom sklearn.model_selection import train_test_split\nimport numpy as np\n\ndef prepare_data():\n # \ub370\uc774\ud130 \uc900\ube44 \ubc0f \ud2b9\uc131 \uc0dd\uc131\n # ... (\ub370\uc774\ud130 \uc218\uc9d1 \ubc0f \uc804\ucc98\ub9ac \uad6c\uac04)\n return X, y\n\ndef initialize(context):\n context.asset = symbol('AAPL')\n context.model = RandomForestClassifier()\n \n X, y = prepare_data()\n context.model.fit(X, y)\n\ndef handle_data(context, data):\n # \ud2b9\uc131 \uc0dd\uc131 \ubc0f \uc608\uce21\n # ... (\ud2b9\uc131 \uc0dd\uc131 \ub85c\uc9c1)\n \n prediction = context.model.predict(X_new)\n if prediction == 1: # \ub9e4\uc218 \uc2e0\ud638\n order(context.asset, 1)\n elif prediction == -1: # \ub9e4\ub3c4 \uc2e0\ud638\n order(context.asset, -1)\n<\/code><\/pre>\n6. \ub9c8\ubb34\ub9ac \ubc0f \uc55e\uc73c\ub85c\uc758 \ubc29\ud5a5<\/h2>\n