\ucd5c\uadfc \uba87 \ub144\uac04 \uae08\uc735 \uc2dc\uc7a5\uc5d0\uc11c \ub370\uc774\ud130 \uae30\ubc18 \uac70\ub798 \ubc29\uc2dd\uc744 \uad6c\ud604\ud558\uae30 \uc704\ud574 \uba38\uc2e0\ub7ec\ub2dd\uacfc \ub525\ub7ec\ub2dd \uc54c\uace0\ub9ac\uc998\uc774 \ub110\ub9ac \ucc44\ud0dd\ub418\uace0 \uc788\uc2b5\ub2c8\ub2e4. \uc774\ub7ec\ud55c \uc54c\uace0\ub9ac\uc998\uc740 \ud070 \ub370\uc774\ud130 \uc138\ud2b8\ub97c \ucc98\ub9ac\ud558\uace0, \ud328\ud134\uc744 \ud559\uc2b5\ud558\uba70, \uc608\uce21\uc744 \uc218\ud589\ud558\ub294 \ub370 \uac15\ub825\ud55c \uc131\ub2a5\uc744 \ubc1c\ud718\ud569\ub2c8\ub2e4. \uc774 \uae00\uc5d0\uc11c\ub294 \uba38\uc2e0\ub7ec\ub2dd \ubc0f \ub525\ub7ec\ub2dd\uc744 \uc774\uc6a9\ud55c \uc54c\uace0\ub9ac\uc998 \ud2b8\ub808\uc774\ub529\uc744 \ub2e4\ub8e8\uace0, \uad50\ucc28 \uac80\uc99d\uc744 \ud1b5\ud574 \ubaa8\ub378\uc758 \uc131\ub2a5\uc744 \ud3c9\uac00\ud558\ub294 \ubc29\ubc95\uc744 \ud30c\uc774\uc36c<\/strong>\uc744 \uc0ac\uc6a9\ud558\uc5ec \uad6c\ud604\ud558\ub294 \ubc29\ubc95\uc5d0 \ub300\ud574 \uc790\uc138\ud788 \uc124\uba85\ud558\uaca0\uc2b5\ub2c8\ub2e4.<\/p>\n \uba38\uc2e0\ub7ec\ub2dd \uae30\ubc18\uc758 \uc54c\uace0\ub9ac\uc998 \ud2b8\ub808\uc774\ub529\uc740 \ub370\uc774\ud130\ub97c \ubd84\uc11d\ud558\uace0 \uc2dc\uc7a5 \ud589\ub3d9\uc744 \uc608\uce21\ud558\ub294 \ub370 \uc788\uc5b4 \ubaa8\ub378\uc744 \ud559\uc2b5\uc2dc\ud0a4\ub294 \uacfc\uc815\uc744 \ud3ec\ud568\ud569\ub2c8\ub2e4. \uc774 \uacfc\uc815\uc5d0\uc11c \ud2b9\uc9d5(feature)<\/strong>\uacfc \ub77c\ubca8(label)<\/strong>\uc758 \uc815\uc758\uac00 \uc911\uc694\ud55c\ub370\uc694, \ud2b9\uc9d5\uc740 \uc785\ub825 \ubcc0\uc218, \ub77c\ubca8\uc740 \uc608\uce21\ud558\uace0\uc790 \ud558\ub294 \uacb0\uacfc\ub97c \uc758\ubbf8\ud569\ub2c8\ub2e4. \uc608\ub97c \ub4e4\uc5b4, \uc8fc\uc2dd \uac00\uaca9 \uc608\uce21\uc5d0\uc11c\ub294 \uacfc\uac70 \uc8fc\uac00, \uac70\ub798\ub7c9, \uae08\ub9ac \ub4f1\uc774 \ud2b9\uc9d5\uc774 \ub420 \uc218 \uc788\uace0, \uc608\uce21 \ub300\uc0c1\uc778 \ub2e4\uc74c \ub0a0\uc758 \uc8fc\uac00\uac00 \ub77c\ubca8\uc774 \ub420 \uc218 \uc788\uc2b5\ub2c8\ub2e4.<\/p>\n \uc54c\uace0\ub9ac\uc998 \ud2b8\ub808\uc774\ub529\uc758 \uccab \ubc88\uc9f8 \ub2e8\uacc4\ub294 \ub370\uc774\ud130 \uc218\uc9d1\uc785\ub2c8\ub2e4. Yahoo Finance, Alpha Vantage \ub4f1 \ub2e4\uc591\ud55c API\uc5d0\uc11c \uc8fc\uac00 \ub370\uc774\ud130\ub97c \uc218\uc9d1\ud560 \uc218 \uc788\uc2b5\ub2c8\ub2e4. \uc218\uc9d1\ud55c \ub370\uc774\ud130\ub294 \uc77c\ubc18\uc801\uc73c\ub85c \ub2e4\uc74c\uacfc \uac19\uc740 \uc804\ucc98\ub9ac \ub2e8\uacc4\ub97c \uac70\uce69\ub2c8\ub2e4:<\/p>\n \ubaa8\ub378\uc744 \uc120\ud0dd\ud560 \ub54c\ub294 \ubb38\uc81c\uc758 \ud2b9\uc131\uacfc \ub370\uc774\ud130\uc758 \uc131\uaca9\uc5d0 \ub530\ub77c \uc801\uc808\ud55c \uc54c\uace0\ub9ac\uc998\uc744 \uc120\ud0dd\ud558\ub294 \uac83\uc774 \uc911\uc694\ud569\ub2c8\ub2e4. \uc8fc\ub85c \uc0ac\uc6a9\ub418\ub294 \uba38\uc2e0\ub7ec\ub2dd \uc54c\uace0\ub9ac\uc998\uc740 \ub2e4\uc74c\uacfc \uac19\uc2b5\ub2c8\ub2e4:<\/p>\n \ubaa8\ub378\uc740 \ub370\uc774\ud130\ub97c \uc0ac\uc6a9\ud558\uc5ec \ub9e4\uac1c\ubcc0\uc218\ub97c \ud559\uc2b5\ud569\ub2c8\ub2e4. \ud30c\uc774\uc36c\uc758 \ubaa8\ub378\uc758 \uc131\ub2a5\uc744 \ud3c9\uac00\ud560 \ub54c \uad50\ucc28 \uac80\uc99d\uc740 \ud544\uc218\uc801\uc785\ub2c8\ub2e4. \uad50\ucc28 \uac80\uc99d\uc758 \ubaa9\uc801\uc740 \uacfc\uc801\ud569(overfitting)\uc744 \ubc29\uc9c0\ud558\uace0, \ubaa8\ub378\uc758 \uc77c\ubc18\ud654 \ub2a5\ub825\uc744 \ub192\uc774\uae30 \uc704\ud568\uc785\ub2c8\ub2e4. \uc77c\ubc18\uc801\uc73c\ub85c K-\uacb9 \uad50\ucc28 \uac80\uc99d<\/strong>\uc774 \ub9ce\uc774 \uc0ac\uc6a9\ub429\ub2c8\ub2e4.<\/p>\n K-\uacb9 \uad50\ucc28 \uac80\uc99d\uc740 \ub370\uc774\ud130\ub97c K\uac1c\uc758 \ubd80\ubd84\uc73c\ub85c \ub098\ub204\uace0, K\ubc88\uc758 \ud559\uc2b5 \ubc0f \uac80\uc99d \uacfc\uc815\uc744 \ud1b5\ud574 \ubaa8\ub378\uc758 \ud3c9\uade0 \uc131\ub2a5\uc744 \ud3c9\uac00\ud558\ub294 \ubc29\ubc95\uc785\ub2c8\ub2e4. \uc608\ub97c \ub4e4\uc5b4, K\uac00 5\uc77c \ub54c, \uc804\uccb4 \ub370\uc774\ud130\ub97c 5\uac1c\uc758 \ud3f4\ub4dc\ub85c \ub098\ub204\uc5b4 4\uac1c\uc758 \ud3f4\ub4dc\ub85c \ud559\uc2b5\ud558\uace0 \ub098\uba38\uc9c0 1\uac1c \ud3f4\ub4dc\ub85c \uac80\uc99d\uc744 \uc218\ud589\ud569\ub2c8\ub2e4.<\/p>\n \uad50\ucc28 \uac80\uc99d\uc744 \uad6c\ud604\ud558\ub294 \ubc29\ubc95\uc740 \uac04\ub2e8\ud558\uba70 \ub525\ub7ec\ub2dd\uc740 \ub354\uc6b1 \ubcf5\uc7a1\ud55c \ub370\uc774\ud130\ub97c \ub2e4\ub8f0 \ub54c \ub6f0\uc5b4\ub09c \uc131\ub2a5\uc744 \ubc1c\ud718\ud569\ub2c8\ub2e4. \ub2e4\uc74c\uc740 Keras\ub97c \uc774\uc6a9\ud55c \uac04\ub2e8\ud55c \ub525\ub7ec\ub2dd \ubaa8\ub378\uc758 \uad6c\ud604 \uc608\uc2dc\uc785\ub2c8\ub2e4:<\/p>\n \uc774 \uae00\uc5d0\uc11c\ub294 \uba38\uc2e0\ub7ec\ub2dd \ubc0f \ub525\ub7ec\ub2dd\uc744 \uae30\ubc18\uc73c\ub85c \ud55c \uc54c\uace0\ub9ac\uc998 \ud2b8\ub808\uc774\ub529\uc758 \uac1c\ub150, \ub370\uc774\ud130 \uc804\ucc98\ub9ac \ubc0f \ubaa8\ub378 \ud559\uc2b5, \uad50\ucc28 \uac80\uc99d\uc758 \ud544\uc694\uc131\uacfc \ud30c\uc774\uc36c\uc744 \ud1b5\ud55c \uad6c\ud604 \ubc29\ubc95\uc5d0 \ub300\ud574 \ub17c\uc758\ud588\uc2b5\ub2c8\ub2e4. \uad50\ucc28 \uac80\uc99d\uc744 \ud1b5\ud574 \ubaa8\ub378\uc758 \uc77c\ubc18\ud654 \ub2a5\ub825\uc744 \ub192\uc774\ub294 \uacfc\uc815\uc740 \uc2e0\ub8b0\ud560 \uc218 \uc788\ub294 \ud2b8\ub808\uc774\ub529 \uc804\ub7b5\uc744 \uc218\ub9bd\ud558\ub294 \ub370 \ud544\uc218\uc801\uc785\ub2c8\ub2e4. \uc774\ub7ec\ud55c \uae30\uc220\uc744 \ubc14\ud0d5\uc73c\ub85c \ub370\uc774\ud130 \uae30\ubc18\uc758 \ud6a8\uc728\uc801\uc774\uace0 \uc218\uc775\uc131 \uc788\ub294 \uac70\ub798 \uc804\ub7b5 \uac1c\ubc1c\uc774 \uac00\ub2a5\ud574\uc9d1\ub2c8\ub2e4.<\/p>\n \uc774\uc81c \uc5ec\ub7ec\ubd84\ub3c4 Python\uacfc \ub2e4\uc591\ud55c \uba38\uc2e0\ub7ec\ub2dd\/\ub525\ub7ec\ub2dd \ub77c\uc774\ube0c\ub7ec\ub9ac\ub97c \uc0ac\uc6a9\ud558\uc5ec \uc790\uc2e0\ub9cc\uc758 \uc54c\uace0\ub9ac\uc998 \ud2b8\ub808\uc774\ub529 \uc804\ub7b5\uc744 \uad6c\uc0c1\ud558\uace0, \uad50\ucc28 \uac80\uc99d\uc744 \ud1b5\ud574 \ubaa8\ub378\uc758 \uc2e0\ub8b0\uc131\uc744 \ud5a5\uc0c1\uc2dc\ud0ac \uc218 \uc788\uc2b5\ub2c8\ub2e4. \uae08\uc735 \uc2dc\uc7a5\uc5d0\uc11c \uc131\uacf5\uc801\uc778 \ud2b8\ub808\uc774\ub529\uc744 \uc704\ud574 \uacc4\uc18d\ud574\uc11c \ud559\uc2b5\ud558\uace0 \uc2e4\ud5d8\ud574\ubcf4\uae38 \ubc14\ub78d\ub2c8\ub2e4.<\/p>\n <\/body><\/p>\n","protected":false},"excerpt":{"rendered":"1. \uba38\uc2e0\ub7ec\ub2dd \ubc0f \ub525\ub7ec\ub2dd \uc54c\uace0\ub9ac\uc998 \ud2b8\ub808\uc774\ub529\uc758 \uae30\ubcf8 \uac1c\ub150<\/h2>\n
1.1 \ub370\uc774\ud130 \uc218\uc9d1 \ubc0f \uc804\ucc98\ub9ac<\/h3>\n
\n
2. \uba38\uc2e0\ub7ec\ub2dd \ubaa8\ub378 \uc120\ud0dd \ubc0f \ud559\uc2b5<\/h2>\n
\n
2.1 \ubaa8\ub378 \ud559\uc2b5<\/h3>\n
scikit-learn<\/code> \ub610\ub294 TensorFlow<\/code>, Keras<\/code>\uc640 \uac19\uc740 \ub77c\uc774\ube0c\ub7ec\ub9ac\ub97c \uc0ac\uc6a9\ud558\uc5ec \uac04\ud3b8\ud558\uac8c \uad6c\ud604\ud560 \uc218 \uc788\uc2b5\ub2c8\ub2e4.<\/p>\nfrom sklearn.model_selection import train_test_split\nfrom sklearn.ensemble import RandomForestClassifier\nfrom sklearn.metrics import accuracy_score\n\n# \ub370\uc774\ud130 \ub85c\ub4dc\nX, y = load_data() # \uc0ac\uc6a9\uc790 \uc815\uc758 \ud568\uc218\ub85c \ub370\uc774\ud130 \ub85c\ub4dc\n\n# \ub370\uc774\ud130 \ubd84\ud560\nX_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.2, random_state=42)\n\n# \ubaa8\ub378 \ucd08\uae30\ud654 \ubc0f \ud559\uc2b5\nmodel = RandomForestClassifier(n_estimators=100)\nmodel.fit(X_train, y_train)\n\n# \uc608\uce21 \ubc0f \uc131\ub2a5 \ud3c9\uac00\ny_pred = model.predict(X_test)\naccuracy = accuracy_score(y_test, y_pred)\nprint(\"\uc815\ud655\ub3c4:\", accuracy)<\/code><\/pre>\n3. \uad50\ucc28 \uac80\uc99d\uc758 \ud544\uc694\uc131<\/h2>\n
3.1 K-\uacb9 \uad50\ucc28 \uac80\uc99d<\/h3>\n
4. \ud30c\uc774\uc36c\uc73c\ub85c \uad50\ucc28 \uac80\uc99d \uad6c\ud604\ud558\uae30<\/h2>\n
scikit-learn<\/code> \ub77c\uc774\ube0c\ub7ec\ub9ac\ub97c \ud65c\uc6a9\ud558\uba74 \ub9e4\uc6b0 \ud6a8\uc728\uc801\uc73c\ub85c \uc218\ud589\ud560 \uc218 \uc788\uc2b5\ub2c8\ub2e4. \ub2e4\uc74c\uc740 K-\uacb9 \uad50\ucc28 \uac80\uc99d\uc744 \ud1b5\ud574 \ubaa8\ub378\uc758 \uc131\ub2a5\uc744 \ud3c9\uac00\ud558\ub294 \uacfc\uc815\uc785\ub2c8\ub2e4:<\/p>\nfrom sklearn.model_selection import cross_val_score\n\n# K-\uacb9 \uad50\ucc28 \uac80\uc99d\nscores = cross_val_score(model, X, y, cv=5)\nprint(\"\uad50\ucc28 \uac80\uc99d \uc810\uc218:\", scores)\nprint(\"\ud3c9\uade0 \uc815\ud655\ub3c4:\", scores.mean())<\/code><\/pre>\n5. \ub525\ub7ec\ub2dd \ubaa8\ub378\uc744 \uc774\uc6a9\ud55c \uc5f0\uacb0 \ubc0f \ud3c9\uac00<\/h2>\n
import numpy as np\nfrom keras.models import Sequential\nfrom keras.layers import Dense\n\n# \ubaa8\ub378 \uc0dd\uc131\nmodel = Sequential()\nmodel.add(Dense(64, input_dim=X.shape[1], activation='relu'))\nmodel.add(Dense(32, activation='relu'))\nmodel.add(Dense(1, activation='sigmoid'))\n\n# \ubaa8\ub378 \ucef4\ud30c\uc77c\nmodel.compile(loss='binary_crossentropy', optimizer='adam', metrics=['accuracy'])\n\n# \ubaa8\ub378 \ud6c8\ub828\nmodel.fit(X_train, y_train, epochs=100, batch_size=10, verbose=0)\n\n# \ubaa8\ub378 \ud3c9\uac00\nscores = model.evaluate(X_test, y_test)\nprint(\"\ud14c\uc2a4\ud2b8 \uc815\ud655\ub3c4:\", scores[1])<\/code><\/pre>\n6. \uacb0\ub860<\/h2>\n