\ucd5c\uadfc \uba87 \ub144 \uac04 \ube44\ud2b8\ucf54\uc778\uacfc \uac19\uc740 \uc554\ud638\ud654\ud3d0\uc758 \uc778\uae30\uac00 \uae09\uc99d\ud568\uc5d0 \ub530\ub77c, \ub9ce\uc740 \uac70\ub798\uc790\ub4e4\uc774 \uc790\ub3d9\ub9e4\ub9e4 \uc2dc\uc2a4\ud15c\uc744 \uad6c\ucd95\ud558\uc5ec \uc218\uc775\uc744 \uadf9\ub300\ud654\ud558\uace0\uc790 \ub178\ub825\ud558\uace0 \uc788\uc2b5\ub2c8\ub2e4. \uc774 \uac15\uc88c\uc5d0\uc11c\ub294 \uba38\uc2e0\ub7ec\ub2dd\uc758 \ud55c \uae30\ubc95\uc778 \ub79c\ub364 \ud3ec\ub808\uc2a4\ud2b8(Random Forest)\ub97c \uc0ac\uc6a9\ud558\uc5ec \ube44\ud2b8\ucf54\uc778\uc758 \ub9e4\uc218 \ubc0f \ub9e4\ub3c4 \uc2e0\ud638\ub97c \uc608\uce21\ud558\ub294 \ubc29\ubc95\uc744 \ud559\uc2b5\ud560 \uac83\uc785\ub2c8\ub2e4.<\/p>\n
\ub79c\ub364 \ud3ec\ub808\uc2a4\ud2b8\ub294 \uc5ec\ub7ec \uac1c\uc758 \uacb0\uc815 \ud2b8\ub9ac(Decision Tree)\ub97c \uacb0\ud569\ud558\uc5ec \uc608\uce21\uc744 \uc218\ud589\ud558\ub294 \uc559\uc0c1\ube14 \ud559\uc2b5 \uc54c\uace0\ub9ac\uc998\uc785\ub2c8\ub2e4. \uc774 \uae30\ubc95\uc740 \uc8fc\uc5b4\uc9c4 \ub370\uc774\ud130\uc14b\uc5d0\uc11c \ubb34\uc791\uc704\ub85c \uc0d8\ud50c\ub9c1\ud55c \ub370\uc774\ud130\ub85c \uc5ec\ub7ec \uac1c\uc758 \uacb0\uc815 \ud2b8\ub9ac\ub97c \uc0dd\uc131\ud558\uace0, \uac01 \ud2b8\ub9ac\uac00 \uc608\uce21\ud55c \uac12\uc744 \ud1b5\ud569\ud558\uc5ec \ucd5c\uc885 \uc608\uce21 \uacb0\uacfc\ub97c \uc0dd\uc131\ud569\ub2c8\ub2e4. \ub79c\ub364 \ud3ec\ub808\uc2a4\ud2b8\ub294 \uace0\ucc28\uc6d0 \ub370\uc774\ud130\uc640 \ub178\uc774\uc988\uc5d0 \uac15\ud55c \ud2b9\uc131\uc744 \uac00\uc9c0\uace0 \uc788\uc5b4 \uae08\uc735 \ub370\uc774\ud130\uc758 \uc608\uce21\uc5d0 \uc801\ud569\ud569\ub2c8\ub2e4.<\/p>\n
\ub79c\ub364 \ud3ec\ub808\uc2a4\ud2b8 \ubaa8\ub378\uc744 \ud6c8\ub828\uc2dc\ud0a4\uae30 \uc704\ud574 \ud544\uc694\ud55c \ub370\uc774\ud130\ub294 \ube44\ud2b8\ucf54\uc778\uc758 \uac00\uaca9 \ub370\uc774\ud130\uc640 \uac70\ub798\ub7c9, \uc774\ub3d9 \ud3c9\uade0 \ub4f1 \ub2e4\uc591\ud55c \uc9c0\ud45c\ub4e4\uc785\ub2c8\ub2e4. \ub370\uc774\ud130\ub294 \ub2e4\uc74c\uacfc \uac19\uc740 \ud615\uc2dd\uc73c\ub85c \uc900\ube44\ub418\uc5b4\uc57c \ud569\ub2c8\ub2e4.<\/p>\n
\ub0a0\uc9dc, \uc2dc\uac00, \uace0\uac00, \uc800\uac00, \uc885\uac00, \uac70\ub798\ub7c9\n2021-01-01, 30000, 31000, 29000, 30500, 1000\n2021-01-02, 30500, 31500, 29500, 30000, 850\n...\n<\/code><\/pre>\n2.1 \ub370\uc774\ud130\uc14b \uc218\uc9d1<\/h3>\n
\ube44\ud2b8\ucf54\uc778 \uac00\uaca9 \ub370\uc774\ud130\ub294 \ub2e4\uc591\ud55c \ubc29\ubc95\uc73c\ub85c \uc218\uc9d1\ud560 \uc218 \uc788\uc2b5\ub2c8\ub2e4. API\ub97c \uc0ac\uc6a9\ud558\uc5ec \uc790\ub3d9\uc73c\ub85c \ub370\uc774\ud130\ub97c \uac00\uc838\uc624\uac70\ub098, CSV \ud30c\uc77c\ub85c \ub2e4\uc6b4\ub85c\ub4dc\ud560 \uc218 \uc788\uc2b5\ub2c8\ub2e4. \uc774\ubc88 \uc608\uc81c\uc5d0\uc11c\ub294 Pandas<\/strong> \ub77c\uc774\ube0c\ub7ec\ub9ac\ub97c \uc0ac\uc6a9\ud558\uc5ec CSV \ud30c\uc77c\uc744 \uc77d\ub294 \ubc29\uc2dd\uc744 \ubcf4\uc5ec\uc8fc\uaca0\uc2b5\ub2c8\ub2e4.<\/p>\n2.2 \ub370\uc774\ud130 \uc804\ucc98\ub9ac<\/h3>\nimport pandas as pd\n\n# \ub370\uc774\ud130 \uc77d\uae30\ndata = pd.read_csv('bitcoin_data.csv')\n\n# \ub0a0\uc9dc\ub97c datetime \ud615\uc2dd\uc73c\ub85c \ubcc0\ud658\ndata['\ub0a0\uc9dc'] = pd.to_datetime(data['\ub0a0\uc9dc'])\n\n# \uacb0\uce21\uce58 \ucc98\ub9ac\ndata.fillna(method='ffill', inplace=True)\n<\/code><\/pre>\n3. \ud2b9\uc9d5 \uc5d4\uc9c0\ub2c8\uc5b4\ub9c1<\/h2>\n
\ub79c\ub364 \ud3ec\ub808\uc2a4\ud2b8 \ubaa8\ub378\uc758 \uc131\ub2a5\uc744 \ub192\uc774\uae30 \uc704\ud574\uc11c\ub294 \uc801\uc808\ud55c \ud2b9\uc9d5(feature)\uc744 \uc120\ud0dd\ud558\uace0 \uc0dd\uc131\ud558\ub294 \uac83\uc774 \uc911\uc694\ud569\ub2c8\ub2e4. \ube44\ud2b8\ucf54\uc778\uc758 \uac00\uaca9 \ub370\uc774\ud130\uc5d0\uc11c \uba87 \uac00\uc9c0 \uc911\uc694\ud55c \ud2b9\uc9d5\uc744 \uc0dd\uc131\ud574 \ubcf4\uaca0\uc2b5\ub2c8\ub2e4.<\/p>\n
3.1 \uc774\ub3d9 \ud3c9\uade0(Moving Average)<\/h3>\n
\uac00\uc7a5 \ub2e8\uc21c\ud558\uba74\uc11c\ub3c4 \uc720\uc6a9\ud55c \uc9c0\ud45c \uc911 \ud558\ub098\uc778 \uc774\ub3d9 \ud3c9\uade0\uc744 \uacc4\uc0b0\ud558\uc5ec \ucd94\uac00 \ud2b9\uc131\uc73c\ub85c \uc0ac\uc6a9\ud569\ub2c8\ub2e4.<\/p>\n
# 5\uc77c \uc774\ub3d9 \ud3c9\uade0\ndata['MA5'] = data['\uc885\uac00'].rolling(window=5).mean()\n\n# 10\uc77c \uc774\ub3d9 \ud3c9\uade0\ndata['MA10'] = data['\uc885\uac00'].rolling(window=10).mean()\n<\/code><\/pre>\n3.2 \ubcc0\ub3d9\uc131(Volatility)<\/h3>\n
\ubcc0\ub3d9\uc131\uc740 \uc790\uc0b0\uc758 \uac00\uaca9\uc774 \uc5bc\ub9c8\ub098 \ubcc0\ub3d9\ud558\ub294\uc9c0\ub97c \ub098\ud0c0\ub0b4\ub294 \uc9c0\ud45c\uc785\ub2c8\ub2e4. \uc774\ub97c \uacc4\uc0b0\ud558\uc5ec \ubaa8\ub378\uc758 \uc785\ub825\uc73c\ub85c \uc0ac\uc6a9\ud560 \uc218 \uc788\uc2b5\ub2c8\ub2e4.<\/p>\n
# \ud45c\uc900\ud3b8\ucc28\ub97c \uc0ac\uc6a9\ud558\uc5ec 5\uc77c \ubcc0\ub3d9\uc131 \uacc4\uc0b0\ndata['Volatility'] = data['\uc885\uac00'].rolling(window=5).std()\n<\/code><\/pre>\n4. \ub9e4\uc218\/\ub9e4\ub3c4 \uc2e0\ud638 \uc0dd\uc131<\/h2>\n
\ub9e4\uc218\/\ub9e4\ub3c4 \uc2e0\ud638\ub97c \uc0dd\uc131\ud558\uae30 \uc704\ud574\uc11c\ub294 \uc774\uc804 \ub370\uc774\ud130\uc758 \ud2b9\uc9d5\uc744 \uc0ac\uc6a9\ud558\uc5ec \ubbf8\ub798 \uac00\uaca9 \ubc29\ud5a5\uc131\uc744 \uc608\uce21\ud574\uc57c \ud569\ub2c8\ub2e4. \uc774 \uc608\uc81c\uc5d0\uc11c\ub294 \uc885\uac00 \uc99d\uac00 \uc5ec\ubd80\ub85c \ub9e4\uc218\/\ub9e4\ub3c4 \uc2e0\ud638\ub97c \uc0dd\uc131\ud569\ub2c8\ub2e4.<\/p>\n
data['\uc2e0\ud638'] = 0\ndata.loc[data['\uc885\uac00'].shift(-1) > data['\uc885\uac00'], '\uc2e0\ud638'] = -1 # \ub9e4\ub3c4 \uc2e0\ud638\ndata.loc[data['\uc885\uac00'].shift(-1) < data['\uc885\uac00'], '\uc2e0\ud638'] = 1 # \ub9e4\uc218 \uc2e0\ud638\n<\/code><\/pre>\n4.1 \ud6c8\ub828 \ubc0f \ud14c\uc2a4\ud2b8 \ub370\uc774\ud130 \ubd84\ub9ac<\/h3>\n
\ubaa8\ub378\uc758 \uc131\ub2a5\uc744 \ud3c9\uac00\ud558\uae30 \uc704\ud574 \ud6c8\ub828 \ub370\uc774\ud130\uc640 \ud14c\uc2a4\ud2b8 \ub370\uc774\ud130\ub97c \ub098\ub204\ub3c4\ub85d \ud558\uaca0\uc2b5\ub2c8\ub2e4.<\/p>\n
from sklearn.model_selection import train_test_split\n\n# \ud2b9\uc131\uacfc \ud0c0\uac9f \ubcc0\uc218 \uc815\uc758\nX = data[['MA5', 'MA10', 'Volatility']].iloc[:-1] # \ub9c8\uc9c0\ub9c9 \ud589\uc740 \uc81c\uc678\ny = data['\uc2e0\ud638'].iloc[:-1] # \ub9c8\uc9c0\ub9c9 \ud589\uc740 \uc81c\uc678\n\n# \ud6c8\ub828 \ub370\uc774\ud130\uc640 \ud14c\uc2a4\ud2b8 \ub370\uc774\ud130\ub85c \ubd84\ub9ac\nX_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.2, random_state=42)\n<\/code><\/pre>\n5. \ub79c\ub364 \ud3ec\ub808\uc2a4\ud2b8 \ubaa8\ub378 \ud6c8\ub828<\/h2>\n
\uc774\uc81c \ub79c\ub364 \ud3ec\ub808\uc2a4\ud2b8 \ubaa8\ub378\uc744 \ud6c8\ub828\uc2dc\ud0a4\uace0, \ud14c\uc2a4\ud2b8 \ub370\uc774\ud130\ub97c \uc0ac\uc6a9\ud558\uc5ec \uc608\uce21\uc744 \uc218\ud589\ud574 \ubcf4\uaca0\uc2b5\ub2c8\ub2e4.<\/p>\n
from sklearn.ensemble import RandomForestClassifier\nfrom sklearn.metrics import classification_report, accuracy_score\n\n# \ub79c\ub364 \ud3ec\ub808\uc2a4\ud2b8 \ubaa8\ub378 \ucd08\uae30\ud654\nmodel = RandomForestClassifier(n_estimators=100, random_state=42)\n\n# \ubaa8\ub378 \ud6c8\ub828\nmodel.fit(X_train, y_train)\n\n# \ud14c\uc2a4\ud2b8 \ub370\uc774\ud130\ub85c \uc608\uce21\ny_pred = model.predict(X_test)\n\n# \uc131\ub2a5 \ud3c9\uac00\nprint(\"\uc815\ud655\ub3c4:\", accuracy_score(y_test, y_pred))\nprint(classification_report(y_test, y_pred))\n<\/code><\/pre>\n6. \ub9e4\ub9e4 \uc804\ub7b5 \uac1c\ubc1c<\/h2>\n
\uc608\uce21\ub41c \ub9e4\uc218\/\ub9e4\ub3c4 \uc2e0\ud638\ub97c \ubc14\ud0d5\uc73c\ub85c \ub9e4\ub9e4 \uc804\ub7b5\uc744 \uac1c\ubc1c\ud560 \uc218 \uc788\uc2b5\ub2c8\ub2e4. \uc608\ub97c \ub4e4\uc5b4, \uc608\uce21\ub41c \uc2e0\ud638\uc5d0 \ub530\ub77c \ub9e4\uc218 \ub610\ub294 \ub9e4\ub3c4\ub97c \uc2e4\ud589\ud558\ub294 \uac04\ub2e8\ud55c \uc804\ub7b5\uc744 \uad6c\ud604\ud574 \ubcf4\uaca0\uc2b5\ub2c8\ub2e4.<\/p>\n
def trading_strategy(data, signals):\n cash = 10000 # \ucd08\uae30 \uc790\ubcf8 \uae08\uc561\n position = 0 # \ubcf4\uc720\uc911\uc778 \ube44\ud2b8\ucf54\uc778 \uc218\n for i in range(len(signals)):\n if signals[i] == 1: # \ub9e4\uc218 \uc2e0\ud638\n position += cash \/ data['\uc885\uac00'].iloc[i]\n cash = 0\n elif signals[i] == -1 and position > 0: # \ub9e4\ub3c4 \uc2e0\ud638\n cash += position * data['\uc885\uac00'].iloc[i]\n position = 0\n return cash # \ucd5c\uc885 \uc790\ud3ec\ubd80\n\nfinal_amount = trading_strategy(data.iloc[len(data) - len(y_pred):], y_pred)\nprint(\"\ucd5c\uc885 \uc790\ubcf8 \uae08\uc561:\", final_amount)\n<\/code><\/pre>\n7. \uacb0\ub860 \ubc0f \ubc1c\uc804 \ubc29\ud5a5<\/h2>\n
\uc774\ubc88 \uac15\uc88c\uc5d0\uc11c\ub294 \ub79c\ub364 \ud3ec\ub808\uc2a4\ud2b8\ub97c \uc774\uc6a9\ud558\uc5ec \ube44\ud2b8\ucf54\uc778\uc758 \ub9e4\uc218 \ubc0f \ub9e4\ub3c4 \uc2e0\ud638\ub97c \uc608\uce21\ud558\ub294 \ubc29\ubc95\uc744 \ubc30\uc6e0\uc2b5\ub2c8\ub2e4. \uc774 \uacfc\uc815\uc5d0\uc11c \ub370\uc774\ud130 \uc218\uc9d1, \uc804\ucc98\ub9ac, \ud2b9\uc9d5 \uc5d4\uc9c0\ub2c8\uc5b4\ub9c1, \ubaa8\ub378 \ud6c8\ub828 \ubc0f \ub9e4\ub9e4 \uc804\ub7b5 \uac1c\ubc1c\uae4c\uc9c0\uc758 \uc804\ubc18\uc801\uc778 \uacfc\uc815\uc744 \uc54c\uc544\ubcf4\uc558\uc2b5\ub2c8\ub2e4. \uc55e\uc73c\ub85c\ub294 \ucd94\uac00\uc801\uc778 \uc9c0\ud45c\ub098 \uc2e0\ud638\ub97c \ud1b5\ud55c \uc131\ub2a5 \uac1c\uc120, \ud558\uc774\ud37c \ud30c\ub77c\ubbf8\ud130 \ud29c\ub2dd, \uba38\uc2e0\ub7ec\ub2dd \ubaa8\ub378\uc758 \ud1b5\ud569 \ub4f1 \ub2e4\uc591\ud55c \ubc29\ud5a5\uc73c\ub85c \uc5f0\uad6c\ub97c \uc9c4\ud589\ud560 \uc218 \uc788\uc2b5\ub2c8\ub2e4.<\/p>\n
\ube44\ud2b8\ucf54\uc778 \uc2dc\uc7a5\uc740 \ubcf8\uc9c8\uc801\uc73c\ub85c \ubcc0\ub3d9\uc131\uc774 \ud06c\uace0 \uc608\uce21\ud558\uae30 \uc5b4\ub824\uc6b4 \ud658\uacbd\uc785\ub2c8\ub2e4. \ub530\ub77c\uc11c \uba38\uc2e0\ub7ec\ub2dd\uc744 \ud65c\uc6a9\ud55c \uc790\ub3d9\ub9e4\ub9e4 \uc2dc\uc2a4\ud15c \uad6c\ucd95 \uc2dc \ud56d\uc0c1 \ub9ac\uc2a4\ud06c \uad00\ub9ac \ubc0f \uc801\uc808\ud55c \uc804\ub7b5\uc744 \uc218\ub9bd\ud574\uc57c \ud55c\ub2e4\ub294 \uc810\uc744 \uc78a\uc9c0 \ub9d0\uc544\uc57c \ud569\ub2c8\ub2e4.<\/p>\n