{"id":477828,"date":"2023-08-09T09:21:11","date_gmt":"2023-08-09T09:21:11","guid":{"rendered":""},"modified":"2023-09-05T11:15:32","modified_gmt":"2023-09-05T11:15:32","slug":"lightgbm","status":"publish","type":"wiki","link":"https:\/\/oneproxy.pro\/kr\/wiki\/lightgbm\/","title":{"rendered":"\ub77c\uc774\ud2b8GBM"},"content":{"rendered":"

LightGBM\uc740 \uadf8\ub798\ub514\uc5b8\ud2b8 \ubd80\uc2a4\ud305\uc744 \uc704\ud574 \uc124\uacc4\ub41c \uac15\ub825\ud558\uace0 \ud6a8\uc728\uc801\uc778 \uc624\ud508 \uc18c\uc2a4 \uae30\uacc4 \ud559\uc2b5 \ub77c\uc774\ube0c\ub7ec\ub9ac\uc785\ub2c8\ub2e4. Microsoft\uc5d0\uc11c \uac1c\ubc1c\ud55c \uc774 \uc81c\ud488\uc740 \ub300\uaddc\ubaa8 \ub370\uc774\ud130 \uc138\ud2b8\ub97c \ucc98\ub9ac\ud558\ub294 \uc18d\ub3c4\uc640 \uace0\uc131\ub2a5\uc73c\ub85c \uc778\ud574 \ub370\uc774\ud130 \uacfc\ud559\uc790\uc640 \uc5f0\uad6c\uc6d0\ub4e4 \uc0ac\uc774\uc5d0\uc11c \uc0c1\ub2f9\ud55c \uc778\uae30\ub97c \uc5bb\uc5c8\uc2b5\ub2c8\ub2e4. LightGBM\uc740 \uc57d\ud55c \ud559\uc2b5\uae30(\uc77c\ubc18\uc801\uc73c\ub85c \uc758\uc0ac\uacb0\uc815 \ud2b8\ub9ac)\ub97c \uacb0\ud569\ud558\uc5ec \uac15\ub825\ud55c \uc608\uce21 \ubaa8\ub378\uc744 \uc0dd\uc131\ud558\ub294 \uae30\uacc4 \ud559\uc2b5 \uae30\uc220\uc778 \uadf8\ub798\ub514\uc5b8\ud2b8 \ubd80\uc2a4\ud305 \ud504\ub808\uc784\uc6cc\ud06c\ub97c \uae30\ubc18\uc73c\ub85c \ud569\ub2c8\ub2e4. \ub6f0\uc5b4\ub09c \uc815\ud655\uc131\uc73c\ub85c \ube45\ub370\uc774\ud130\ub97c \ucc98\ub9ac\ud558\ub294 \ub2a5\ub825 \ub355\ubd84\uc5d0 \uc790\uc5f0\uc5b4 \ucc98\ub9ac, \ucef4\ud4e8\ud130 \ube44\uc804, \uc7ac\ubb34 \ubaa8\ub378\ub9c1 \ub4f1 \ub2e4\uc591\ud55c \uc601\uc5ed\uc5d0\uc11c \uc120\ud638\ub418\ub294 \uc120\ud0dd\uc785\ub2c8\ub2e4.<\/p>\n

LightGBM\uc758 \uae30\uc6d0\uacfc \ucd5c\ucd08 \uc5b8\uae09\uc758 \uc5ed\uc0ac<\/h2>\n

LightGBM\uc740 2017\ub144 Microsoft \uc5f0\uad6c\uc6d0\ub4e4\uc774 "LightGBM: A Highly Efficient Gradient Boosting Decision Tree"\ub77c\ub294 \uc81c\ubaa9\uc758 \ub17c\ubb38\uc5d0\uc11c \ucc98\uc74c \uc18c\uac1c\ud588\uc2b5\ub2c8\ub2e4. \uc774 \ub17c\ubb38\uc758 \uc800\uc790\ub294 Guolin Ke, Qi Meng, Thomas Finley, Taifeng Wang, Wei Chen, Weidong Ma, Qiwei Ye \ubc0f Tie-Yan Liu\uc785\ub2c8\ub2e4. \uc774 \ud68d\uae30\uc801\uc778 \uc5f0\uad6c\ub294 LightGBM\uc744 \uacbd\uc7c1\ub825 \uc788\ub294 \uc815\ud655\ub3c4\ub97c \uc720\uc9c0\ud558\uba74\uc11c \uadf8\ub77c\ub514\uc5b8\ud2b8 \ubd80\uc2a4\ud305 \uc54c\uace0\ub9ac\uc998\uc758 \ud6a8\uc728\uc131\uc744 \ub192\uc774\ub294 \uc0c8\ub85c\uc6b4 \ubc29\ubc95\uc73c\ub85c \uc81c\uc2dc\ud588\uc2b5\ub2c8\ub2e4.<\/p>\n

LightGBM\uc5d0 \ub300\ud55c \uc790\uc138\ud55c \uc815\ubcf4<\/h2>\n

LightGBM\uc740 \uace0\uc720\ud55c \uae30\ub2a5\uc73c\ub85c \uadf8\ub77c\ub514\uc5b8\ud2b8 \ubd80\uc2a4\ud305 \ubd84\uc57c\uc5d0 \ud601\uba85\uc744 \uc77c\uc73c\ucf30\uc2b5\ub2c8\ub2e4. \uae4a\uc774\ubcc4 \ud2b8\ub9ac \uc131\uc7a5\uc744 \uc0ac\uc6a9\ud558\ub294 \uae30\uc874 \uadf8\ub798\ub514\uc5b8\ud2b8 \ubd80\uc2a4\ud305 \ud504\ub808\uc784\uc6cc\ud06c\uc640 \ub2ec\ub9ac LightGBM\uc740 \ub9ac\ud504\ubcc4 \ud2b8\ub9ac \uc131\uc7a5 \uc804\ub7b5\uc744 \uc0ac\uc6a9\ud569\ub2c8\ub2e4. \uc774 \uc811\uadfc \ubc29\uc2dd\uc740 \uac01 \ud2b8\ub9ac \ud655\uc7a5 \uc911\uc5d0 \uc190\uc2e4\uc774 \ucd5c\ub300\ub85c \uac10\uc18c\ud55c \ub9ac\ud504 \ub178\ub4dc\ub97c \uc120\ud0dd\ud558\ubbc0\ub85c \ub354 \uc801\uc740 \uc218\uc758 \ub9ac\ud504\ub85c \ubcf4\ub2e4 \uc815\ud655\ud55c \ubaa8\ub378\uc774 \uc0dd\uc131\ub429\ub2c8\ub2e4.<\/p>\n

\ub610\ud55c LightGBM\uc740 GOSS(Gradient-based One-Side Sampling)\uc640 EFB(Exclusive Feature Bundling)\ub77c\ub294 \ub450 \uac00\uc9c0 \uae30\uc220\uc744 \ud1b5\ud574 \uba54\ubaa8\ub9ac \uc0ac\uc6a9\uc744 \ucd5c\uc801\ud654\ud569\ub2c8\ub2e4. GOSS\ub294 \ud6c8\ub828 \uacfc\uc815\uc5d0\uc11c \uc911\uc694\ud55c \uae30\uc6b8\uae30\ub9cc \uc120\ud0dd\ud558\uc5ec \ubaa8\ub378 \uc815\ud655\ub3c4\ub97c \uc720\uc9c0\ud558\uba74\uc11c \ub370\uc774\ud130 \uc778\uc2a4\ud134\uc2a4 \uc218\ub97c \uc904\uc785\ub2c8\ub2e4. EFB \uadf8\ub8f9\uc5d0\ub294 \uba54\ubaa8\ub9ac\ub97c \uc555\ucd95\ud558\uace0 \ud6a8\uc728\uc131\uc744 \ud5a5\uc0c1\uc2dc\ud0a4\ub294 \ub3c5\uc810 \uae30\ub2a5\uc774 \uc788\uc2b5\ub2c8\ub2e4.<\/p>\n

\ub610\ud55c \ub77c\uc774\ube0c\ub7ec\ub9ac\ub294 \ud68c\uadc0, \ubd84\ub958, \uc21c\uc704 \uc9c0\uc815, \ucd94\ucc9c \uc2dc\uc2a4\ud15c\uacfc \uac19\uc740 \ub2e4\uc591\ud55c \uae30\uacc4 \ud559\uc2b5 \uc791\uc5c5\uc744 \uc9c0\uc6d0\ud569\ub2c8\ub2e4. Python, R, C++\uc640 \uac19\uc740 \uc5ec\ub7ec \ud504\ub85c\uadf8\ub798\ubc0d \uc5b8\uc5b4\ub85c \uc720\uc5f0\ud55c API\ub97c \uc81c\uacf5\ud558\ubbc0\ub85c \ub2e4\uc591\ud55c \ud50c\ub7ab\ud3fc\uc5d0\uc11c \uac1c\ubc1c\uc790\uac00 \uc27d\uac8c \uc561\uc138\uc2a4\ud560 \uc218 \uc788\uc2b5\ub2c8\ub2e4.<\/p>\n

LightGBM\uc758 \ub0b4\ubd80 \uad6c\uc870: LightGBM \uc791\ub3d9 \ubc29\uc2dd<\/h2>\n

LightGBM\uc758 \ud575\uc2ec\uc740 \uc5ec\ub7ec \uc57d\ud55c \ud559\uc2b5\uae30\uac00 \uacb0\ud569\ub418\uc5b4 \uac15\ub825\ud55c \uc608\uce21 \ubaa8\ub378\uc744 \ud615\uc131\ud558\ub294 \uc559\uc0c1\ube14 \ud559\uc2b5 \ubc29\ubc95\uc778 \uadf8\ub798\ub514\uc5b8\ud2b8 \ubd80\uc2a4\ud305 \uae30\uc220\uc744 \uae30\ubc18\uc73c\ub85c \uc791\ub3d9\ud569\ub2c8\ub2e4. LightGBM\uc758 \ub0b4\ubd80 \uad6c\uc870\ub294 \ub2e4\uc74c \ub2e8\uacc4\ub85c \uc694\uc57d\ub420 \uc218 \uc788\uc2b5\ub2c8\ub2e4.<\/p>\n

    \n
  1. \n

    \ub370\uc774\ud130 \uc900\ube44<\/strong>: LightGBM\uc5d0\uc11c\ub294 \uc131\ub2a5\uc744 \ud5a5\uc0c1\ud558\uace0 \uba54\ubaa8\ub9ac \uc0ac\uc6a9\ub7c9\uc744 \uc904\uc774\uae30 \uc704\ud574 Dataset \ub610\ub294 DMatrix\uc640 \uac19\uc740 \ud2b9\uc815 \ud615\uc2dd\uc73c\ub85c \ub370\uc774\ud130\ub97c \uad6c\uc131\ud574\uc57c \ud569\ub2c8\ub2e4.<\/p>\n<\/li>\n

  2. \n

    \ud2b8\ub9ac \uad6c\uc131<\/strong>: \ud6c8\ub828 \uc911\uc5d0 LightGBM\uc740 \ub9ac\ud504 \ubc29\ud5a5 \ud2b8\ub9ac \uc131\uc7a5 \uc804\ub7b5\uc744 \uc0ac\uc6a9\ud569\ub2c8\ub2e4. \ub2e8\uc77c \ub9ac\ud504\ub97c \ub8e8\ud2b8 \ub178\ub4dc\ub85c \uc2dc\uc791\ud55c \ub2e4\uc74c \ub9ac\ud504 \ub178\ub4dc\ub97c \ubd84\ud560\ud558\uc5ec \ubc18\ubcf5\uc801\uc73c\ub85c \ud2b8\ub9ac\ub97c \ud655\uc7a5\ud558\uc5ec \uc190\uc2e4 \ud568\uc218\ub97c \ucd5c\uc18c\ud654\ud569\ub2c8\ub2e4.<\/p>\n<\/li>\n

  3. \n

    \uc78e\ubcc4 \uc131\uc7a5<\/strong>: LightGBM\uc740 \uac00\uc7a5 \ud070 \uc190\uc2e4 \uac10\uc18c\ub97c \uc81c\uacf5\ud558\ub294 \ub9ac\ud504 \ub178\ub4dc\ub97c \uc120\ud0dd\ud558\uc5ec \ub354 \uc801\uc740 \ub9ac\ud504\ub85c \ubcf4\ub2e4 \uc815\ud655\ud55c \ubaa8\ub378\uc744 \uc0dd\uc131\ud569\ub2c8\ub2e4.<\/p>\n<\/li>\n

  4. \n

    GOSS(\uadf8\ub77c\ub514\uc5b8\ud2b8 \uae30\ubc18 \ub2e8\ubc29\ud5a5 \uc0d8\ud50c\ub9c1)<\/strong>: \ud6c8\ub828 \uc911\uc5d0 GOSS\ub294 \ucd94\uac00 \ucd5c\uc801\ud654\ub97c \uc704\ud574 \uc911\uc694\ud55c \uae30\uc6b8\uae30\ub9cc \uc120\ud0dd\ud558\ubbc0\ub85c \uc218\ub834 \uc18d\ub3c4\uac00 \ube68\ub77c\uc9c0\uace0 \uacfc\uc801\ud569\uc774 \uc904\uc5b4\ub4ed\ub2c8\ub2e4.<\/p>\n<\/li>\n

  5. \n

    EFB(\ub3c5\uc810 \uae30\ub2a5 \ubc88\ub4e4\ub9c1)<\/strong>: EFB \uadf8\ub8f9\uc5d0\ub294 \uba54\ubaa8\ub9ac\ub97c \uc808\uc57d\ud558\uace0 \ud6c8\ub828 \uacfc\uc815\uc758 \uc18d\ub3c4\ub97c \ub192\uc774\ub294 \ub3c5\uc810 \uae30\ub2a5\uc774 \uc788\uc2b5\ub2c8\ub2e4.<\/p>\n<\/li>\n

  6. \n

    \ubd80\uc2a4\ud305<\/strong>: \uc57d\ud55c \ud559\uc2b5\uae30(\uc758\uc0ac\uacb0\uc815 \ud2b8\ub9ac)\uac00 \uc21c\ucc28\uc801\uc73c\ub85c \ubaa8\ub378\uc5d0 \ucd94\uac00\ub418\uba70 \uac01\uac01\uc758 \uc0c8 \ud2b8\ub9ac\ub294 \uc774\uc804 \ud2b8\ub9ac\uc758 \uc624\ub958\ub97c \uc218\uc815\ud569\ub2c8\ub2e4.<\/p>\n<\/li>\n

  7. \n

    \uc815\uaddc\ud654<\/strong>: LightGBM\uc740 L1 \ubc0f L2 \uc815\uaddc\ud654 \uae30\uc220\uc744 \uc0ac\uc6a9\ud558\uc5ec \uacfc\uc801\ud569\uc744 \ubc29\uc9c0\ud558\uace0 \uc77c\ubc18\ud654\ub97c \uac1c\uc120\ud569\ub2c8\ub2e4.<\/p>\n<\/li>\n

  8. \n

    \uc608\uce21<\/strong>: \ubaa8\ub378\uc774 \ud6c8\ub828\ub418\uba74 LightGBM\uc740 \uc0c8\ub85c\uc6b4 \ub370\uc774\ud130\uc5d0 \ub300\ud55c \uacb0\uacfc\ub97c \ud6a8\uc728\uc801\uc73c\ub85c \uc608\uce21\ud560 \uc218 \uc788\uc2b5\ub2c8\ub2e4.<\/p>\n<\/li>\n<\/ol>\n

    LightGBM\uc758 \uc8fc\uc694 \uae30\ub2a5 \ubd84\uc11d<\/h2>\n

    LightGBM\uc740 \uad11\ubc94\uc704\ud55c \ucc44\ud0dd\uacfc \ud6a8\uc728\uc131\uc5d0 \uae30\uc5ec\ud558\ub294 \uba87 \uac00\uc9c0 \uc8fc\uc694 \uae30\ub2a5\uc744 \uc790\ub791\ud569\ub2c8\ub2e4.<\/p>\n

      \n
    1. \n

      \uace0\uc18d<\/strong>: \ub9ac\ud504\ubcc4 \ud2b8\ub9ac \uc131\uc7a5 \ubc0f GOSS \ucd5c\uc801\ud654 \uae30\uc220\uc740 LightGBM\uc744 \ub2e4\ub978 \uadf8\ub798\ub514\uc5b8\ud2b8 \ubd80\uc2a4\ud305 \ud504\ub808\uc784\uc6cc\ud06c\ubcf4\ub2e4 \ud6e8\uc52c \ube60\ub974\uac8c \ub9cc\ub4ed\ub2c8\ub2e4.<\/p>\n<\/li>\n

    2. \n

      \uba54\ubaa8\ub9ac \ud6a8\uc728\uc131<\/strong>: EFB \ubc29\ubc95\uc740 \uba54\ubaa8\ub9ac \uc18c\ube44\ub97c \uc904\uc5ec LightGBM\uc774 \uae30\uc874 \uc54c\uace0\ub9ac\uc998\uc744 \uc0ac\uc6a9\ud558\uc5ec \uba54\ubaa8\ub9ac\uc5d0 \ub9de\uc9c0 \uc54a\uc744 \uc218 \uc788\ub294 \ub300\uaddc\ubaa8 \ub370\uc774\ud130 \uc138\ud2b8\ub97c \ucc98\ub9ac\ud560 \uc218 \uc788\ub3c4\ub85d \ud569\ub2c8\ub2e4.<\/p>\n<\/li>\n

    3. \n

      \ud655\uc7a5\uc131<\/strong>: LightGBM\uc740 \uc218\ubc31\ub9cc \uac1c\uc758 \uc778\uc2a4\ud134\uc2a4\uc640 \uae30\ub2a5\uc744 \uac16\ucd98 \ub300\uaddc\ubaa8 \ub370\uc774\ud130\uc138\ud2b8\ub97c \ucc98\ub9ac\ud560 \uc218 \uc788\ub3c4\ub85d \ud6a8\uc728\uc801\uc73c\ub85c \ud655\uc7a5\ub429\ub2c8\ub2e4.<\/p>\n<\/li>\n

    4. \n

      \uc720\uc5f0\uc131<\/strong>: LightGBM\uc740 \ub2e4\uc591\ud55c \uba38\uc2e0\ub7ec\ub2dd \uc791\uc5c5\uc744 \uc9c0\uc6d0\ud558\ubbc0\ub85c \ud68c\uadc0, \ubd84\ub958, \uc21c\uc704 \uc9c0\uc815, \ucd94\ucc9c \uc2dc\uc2a4\ud15c\uc5d0 \uc801\ud569\ud569\ub2c8\ub2e4.<\/p>\n<\/li>\n

    5. \n

      \uc815\ud655\ud55c \uc608\uce21<\/strong>: \uc78e\ubcc4 \ub098\ubb34 \uc131\uc7a5 \uc804\ub7b5\uc740 \ub354 \uc801\uc740 \uc218\uc758 \uc78e\uc744 \uc0ac\uc6a9\ud558\uc5ec \ubaa8\ub378\uc758 \uc608\uce21 \uc815\ud655\ub3c4\ub97c \ud5a5\uc0c1\uc2dc\ud0b5\ub2c8\ub2e4.<\/p>\n<\/li>\n

    6. \n

      \ubc94\uc8fc\ud615 \uae30\ub2a5 \uc9c0\uc6d0<\/strong>: LightGBM\uc740 \uad11\ubc94\uc704\ud55c \uc804\ucc98\ub9ac \uc5c6\uc774\ub3c4 \ubc94\uc8fc\ud615 \uae30\ub2a5\uc744 \ud6a8\uc728\uc801\uc73c\ub85c \ucc98\ub9ac\ud569\ub2c8\ub2e4.<\/p>\n<\/li>\n

    7. \n

      \ubcd1\ub82c \ud559\uc2b5<\/strong>: LightGBM\uc740 \uba40\ud2f0 \ucf54\uc5b4 CPU\ub97c \ud65c\uc6a9\ud558\uc5ec \uc131\ub2a5\uc744 \ub354\uc6b1 \ud5a5\uc0c1\uc2dc\ud0a4\ub294 \ubcd1\ub82c \ud6c8\ub828\uc744 \uc9c0\uc6d0\ud569\ub2c8\ub2e4.<\/p>\n<\/li>\n<\/ol>\n

      LightGBM\uc758 \uc885\ub958<\/h2>\n

      LightGBM\uc740 \uc0ac\uc6a9\ub41c \ubd80\uc2a4\ud305 \uc720\ud615\uc5d0 \ub530\ub77c \ub450 \uac00\uc9c0 \uc8fc\uc694 \uc720\ud615\uc744 \uc81c\uacf5\ud569\ub2c8\ub2e4.<\/p>\n

        \n
      1. \n

        GBM(\uadf8\ub77c\ub514\uc5b8\ud2b8 \ubd80\uc2a4\ud305 \uba38\uc2e0)<\/strong>: \uc774\ub294 \ub9ac\ud504\ubcc4 \ud2b8\ub9ac \uc131\uc7a5 \uc804\ub7b5\uacfc \ud568\uaed8 \uadf8\ub77c\ub514\uc5b8\ud2b8 \ubd80\uc2a4\ud305\uc744 \uc0ac\uc6a9\ud558\ub294 LightGBM\uc758 \ud45c\uc900 \ud615\uc2dd\uc785\ub2c8\ub2e4.<\/p>\n<\/li>\n

      2. \n

        \ub2e4\ud2b8<\/strong>: Dart\ub294 \ud6c8\ub828 \uc911\uc5d0 \ub4dc\ub86d\uc544\uc6c3 \uae30\ubc18 \uc815\uaddc\ud654\ub97c \ud65c\uc6a9\ud558\ub294 LightGBM\uc758 \ubcc0\ud615\uc785\ub2c8\ub2e4. \uac01 \ubc18\ubcf5 \uc911\uc5d0 \uc77c\ubd80 \ub098\ubb34\ub97c \ubb34\uc791\uc704\ub85c \uc0ad\uc81c\ud558\uc5ec \uacfc\uc801\ud569\uc744 \ubc29\uc9c0\ud558\ub294 \ub370 \ub3c4\uc6c0\uc774 \ub429\ub2c8\ub2e4.<\/p>\n<\/li>\n<\/ol>\n

        \ub2e4\uc74c\uc740 GBM\uacfc Dart\uc758 \uc8fc\uc694 \ucc28\uc774\uc810\uc744 \uac15\uc870\ud558\ub294 \ube44\uad50\ud45c\uc785\ub2c8\ub2e4.<\/p>\n\n\n\n\n\n\n\n\n
        \uce21\uba74<\/th>\nGBM(\uadf8\ub77c\ub514\uc5b8\ud2b8 \ubd80\uc2a4\ud305 \uba38\uc2e0)<\/th>\n\ub2e4\ud2b8<\/th>\n<\/tr>\n<\/thead>\n
        \ubd80\uc2a4\ud305 \uc54c\uace0\ub9ac\uc998<\/td>\n\uadf8\ub798\ub514\uc5b8\ud2b8 \ubd80\uc2a4\ud305<\/td>\nDart\ub97c \uc0ac\uc6a9\ud55c \uadf8\ub77c\ub514\uc5b8\ud2b8 \ubd80\uc2a4\ud305<\/td>\n<\/tr>\n
        \uc815\uaddc\ud654 \uae30\ubc95<\/td>\nL1\uacfc L2<\/td>\n\ub4dc\ub86d\uc544\uc6c3\uc774 \uc788\ub294 L1 \ubc0f L2<\/td>\n<\/tr>\n
        \uacfc\uc801\ud569 \ubc29\uc9c0<\/td>\n\ubcf4\ud1b5\uc758<\/td>\n\ub4dc\ub86d\uc544\uc6c3\uc73c\ub85c \uac1c\uc120\ub428<\/td>\n<\/tr>\n
        \ub098\ubb34 \uac00\uc9c0\uce58\uae30<\/td>\n\uac00\uc9c0\uce58\uae30 \uc5c6\uc74c<\/td>\n\ub4dc\ub86d\uc544\uc6c3\uc744 \uae30\ubc18\uc73c\ub85c \ud55c \uac00\uc9c0\uce58\uae30<\/td>\n<\/tr>\n<\/tbody>\n<\/table>\n

        LightGBM \uc0ac\uc6a9\ubc29\ubc95, \uc0ac\uc6a9\uc5d0 \ub530\ub978 \ubb38\uc81c\uc810 \ubc0f \ud574\uacb0\ubc29\ubc95<\/h2>\n

        LightGBM\uc740 \ub2e4\uc591\ud55c \uae30\uacc4 \ud559\uc2b5 \uc791\uc5c5\uc744 \ucc98\ub9ac\ud558\uae30 \uc704\ud574 \ub2e4\uc591\ud55c \ubc29\ubc95\uc73c\ub85c \ud65c\uc6a9\ub420 \uc218 \uc788\uc2b5\ub2c8\ub2e4.<\/p>\n

          \n
        1. \n

          \ubd84\ub958<\/strong>: \uc2a4\ud338 \uac10\uc9c0, \uac10\uc815 \ubd84\uc11d, \uc774\ubbf8\uc9c0 \uc778\uc2dd \ub4f1 \uc774\uc9c4 \ub610\ub294 \ub2e4\uc911 \ud074\ub798\uc2a4 \ubd84\ub958 \ubb38\uc81c\uc5d0 LightGBM\uc744 \uc0ac\uc6a9\ud569\ub2c8\ub2e4.<\/p>\n<\/li>\n

        2. \n

          \ud68c\uadc0<\/strong>: \uc8fc\ud0dd \uac00\uaca9, \uc8fc\uc2dd \uc2dc\uc7a5 \uac00\uce58 \ub610\ub294 \uae30\uc628 \uc608\uce21\uacfc \uac19\uc740 \ud68c\uadc0 \uc791\uc5c5\uc5d0 LightGBM\uc744 \uc801\uc6a9\ud569\ub2c8\ub2e4.<\/p>\n<\/li>\n

        3. \n

          \uc21c\uc704<\/strong>: LightGBM\uc744 \ud65c\uc6a9\ud558\uc5ec \uac80\uc0c9 \uc5d4\uc9c4 \uacb0\uacfc \uc21c\uc704 \ub610\ub294 \ucd94\ucc9c \uc2dc\uc2a4\ud15c\uacfc \uac19\uc740 \uc21c\uc704 \uc2dc\uc2a4\ud15c\uc744 \uad6c\ucd95\ud569\ub2c8\ub2e4.<\/p>\n<\/li>\n

        4. \n

          \ucd94\ucc9c \uc2dc\uc2a4\ud15c<\/strong>: LightGBM\uc740 \uac1c\uc778\ud654\ub41c \ucd94\ucc9c \uc5d4\uc9c4\uc744 \uad6c\ub3d9\ud558\uc5ec \uc0ac\uc6a9\uc790\uc5d0\uac8c \uc81c\ud488, \uc601\ud654 \ub610\ub294 \uc74c\uc545\uc744 \uc81c\uc548\ud560 \uc218 \uc788\uc2b5\ub2c8\ub2e4.<\/p>\n<\/li>\n<\/ol>\n

          \uc7a5\uc810\uc5d0\ub3c4 \ubd88\uad6c\ud558\uace0 \uc0ac\uc6a9\uc790\ub294 LightGBM\uc744 \uc0ac\uc6a9\ud558\ub294 \ub3d9\uc548 \uba87 \uac00\uc9c0 \ubb38\uc81c\uc5d0 \uc9c1\uba74\ud560 \uc218 \uc788\uc2b5\ub2c8\ub2e4.<\/p>\n

            \n
          1. \n

            \ubd88\uade0\ud615 \ub370\uc774\ud130\uc138\ud2b8<\/strong>: LightGBM\uc740 \ubd88\uade0\ud615\ud55c \ub370\uc774\ud130 \uc138\ud2b8\ub85c \uc778\ud574 \ud3b8\ud5a5\ub41c \uc608\uce21\uc73c\ub85c \uc774\uc5b4\uc9c8 \uc218 \uc788\uc2b5\ub2c8\ub2e4. \ud55c \uac00\uc9c0 \ud574\uacb0\ucc45\uc740 \ud074\ub798\uc2a4 \uac00\uc911\uce58 \ub610\ub294 \uc0d8\ud50c\ub9c1 \uae30\uc220\uc744 \uc0ac\uc6a9\ud558\uc5ec \ud6c8\ub828 \uc911\uc5d0 \ub370\uc774\ud130 \uade0\ud615\uc744 \ub9de\ucd94\ub294 \uac83\uc785\ub2c8\ub2e4.<\/p>\n<\/li>\n

          2. \n

            \uacfc\uc801\ud569<\/strong>: LightGBM\uc740 \uacfc\uc801\ud569\uc744 \ubc29\uc9c0\ud558\uae30 \uc704\ud574 \uc815\uaddc\ud654 \uae30\uc220\uc744 \uc0ac\uc6a9\ud558\uc9c0\ub9cc \ub370\uc774\ud130\uac00 \ucda9\ubd84\ud558\uc9c0 \uc54a\uac70\ub098 \ubaa8\ub378\uc774 \ub108\ubb34 \ubcf5\uc7a1\ud55c \uacbd\uc6b0 \uc5ec\uc804\ud788 \ubc1c\uc0dd\ud560 \uc218 \uc788\uc2b5\ub2c8\ub2e4. \uad50\ucc28 \uac80\uc99d \ubc0f \ucd08\ub9e4\uac1c\ubcc0\uc218 \uc870\uc815\uc740 \uc774 \ubb38\uc81c\ub97c \uc644\ud654\ud558\ub294 \ub370 \ub3c4\uc6c0\uc774 \ub420 \uc218 \uc788\uc2b5\ub2c8\ub2e4.<\/p>\n<\/li>\n

          3. \n

            \ucd08\ub9e4\uac1c\ubcc0\uc218 \uc870\uc815<\/strong>: LightGBM\uc758 \uc131\ub2a5\uc740 \ud558\uc774\ud37c\ud30c\ub77c\ubbf8\ud130 \ud29c\ub2dd\uc5d0 \ud06c\uac8c \uc88c\uc6b0\ub429\ub2c8\ub2e4. \ucd5c\uc801\uc758 \ud558\uc774\ud37c\ud30c\ub77c\ubbf8\ud130 \uc870\ud569\uc744 \ucc3e\uae30 \uc704\ud574 \uadf8\ub9ac\ub4dc \uac80\uc0c9\uc774\ub098 \ubca0\uc774\uc9c0\uc548 \ucd5c\uc801\ud654\ub97c \uc0ac\uc6a9\ud560 \uc218 \uc788\uc2b5\ub2c8\ub2e4.<\/p>\n<\/li>\n

          4. \n

            \ub370\uc774\ud130 \uc804\ucc98\ub9ac<\/strong>: \ubc94\uc8fc\ud615 \uae30\ub2a5\uc5d0\ub294 \uc801\uc808\ud55c \uc778\ucf54\ub529\uc774 \ud544\uc694\ud558\uba70 \ub204\ub77d\ub41c \ub370\uc774\ud130\ub294 LightGBM\uc5d0 \uacf5\uae09\ud558\uae30 \uc804\uc5d0 \uc801\uc808\ud558\uac8c \ucc98\ub9ac\ub418\uc5b4\uc57c \ud569\ub2c8\ub2e4.<\/p>\n<\/li>\n<\/ol>\n

            \uc8fc\uc694 \ud2b9\uc9d5 \ubc0f \uae30\ud0c0 \uc720\uc0ac \uc6a9\uc5b4\uc640\uc758 \ube44\uad50<\/h2>\n

            LightGBM\uc744 \ub2e4\ub978 \uc778\uae30 \uc788\ub294 \uadf8\ub798\ub514\uc5b8\ud2b8 \ubd80\uc2a4\ud305 \ub77c\uc774\ube0c\ub7ec\ub9ac\uc640 \ube44\uad50\ud574 \ubcf4\uaca0\uc2b5\ub2c8\ub2e4.<\/p>\n\n\n\n\n\n\n\n\n\n
            \ud2b9\uc131<\/th>\n\ub77c\uc774\ud2b8GBM<\/th>\nXGBoost<\/th>\n\ucea3\ubd80\uc2a4\ud2b8<\/th>\n<\/tr>\n<\/thead>\n
            \ub098\ubb34 \uc131\uc7a5 \uc804\ub7b5<\/td>\n\uc78e \ubc29\ud5a5<\/td>\n\ub808\ubca8\ubcc4<\/td>\n\ub300\uce6d<\/td>\n<\/tr>\n
            \uba54\ubaa8\ub9ac \uc0ac\uc6a9\ub7c9<\/td>\n\ud6a8\uc728\uc801\uc778<\/td>\n\ubcf4\ud1b5\uc758<\/td>\n\ubcf4\ud1b5\uc758<\/td>\n<\/tr>\n
            \ubc94\uc8fc\ud615 \uc9c0\uc6d0<\/td>\n\uc608<\/td>\n\uc81c\ud55c\ub41c<\/td>\n\uc608<\/td>\n<\/tr>\n
            GPU \uac00\uc18d<\/td>\n\uc608<\/td>\n\uc608<\/td>\n\uc81c\ud55c\ub41c<\/td>\n<\/tr>\n
            \uc131\ub2a5<\/td>\n\ub354 \ube60\ub974\uac8c<\/td>\nLGBM\ubcf4\ub2e4 \ub290\ub9bc<\/td>\n\uc720\uc0ac\ud55c<\/td>\n<\/tr>\n<\/tbody>\n<\/table>\n

            LightGBM\uc740 \uc18d\ub3c4 \uce21\uba74\uc5d0\uc11c XGBoost\ubcf4\ub2e4 \uc131\ub2a5\uc774 \ub6f0\uc5b4\ub098\uace0 CatBoost\uc640 LightGBM\uc740 \uc131\ub2a5\uba74\uc5d0\uc11c \ube44\uad50\uc801 \uc720\uc0ac\ud569\ub2c8\ub2e4. LightGBM\uc740 \ub300\uaddc\ubaa8 \ub370\uc774\ud130 \uc138\ud2b8\ub97c \ucc98\ub9ac\ud558\uace0 \uba54\ubaa8\ub9ac\ub97c \ud6a8\uc728\uc801\uc73c\ub85c \ud65c\uc6a9\ud558\ub294 \ub370 \ud0c1\uc6d4\ud558\ubbc0\ub85c \ube45 \ub370\uc774\ud130 \uc2dc\ub098\ub9ac\uc624\uc5d0\uc11c \uc120\ud638\ub418\ub294 \uc120\ud0dd\uc785\ub2c8\ub2e4.<\/p>\n

            LightGBM\uacfc \uad00\ub828\ub41c \ubbf8\ub798 \uc804\ub9dd\uacfc \uae30\uc220<\/h2>\n

            \uae30\uacc4 \ud559\uc2b5 \ubd84\uc57c\uac00 \ubc1c\uc804\ud568\uc5d0 \ub530\ub77c LightGBM\uc740 \ub354\uc6b1 \uac1c\uc120\ub418\uace0 \ubc1c\uc804\ud560 \uac00\ub2a5\uc131\uc774 \ub192\uc2b5\ub2c8\ub2e4. \uc7a0\uc7ac\uc801\uc778 \ubbf8\ub798 \uac1c\ubc1c\uc5d0\ub294 \ub2e4\uc74c\uc774 \ud3ec\ud568\ub429\ub2c8\ub2e4.<\/p>\n

              \n
            1. \n

              \ud5a5\uc0c1\ub41c \uc815\uaddc\ud654 \uae30\uc220<\/strong>: \uc5f0\uad6c\uc790\ub4e4\uc740 \ubcf5\uc7a1\ud55c \ub370\uc774\ud130 \uc138\ud2b8\ub97c \uc77c\ubc18\ud654\ud558\uace0 \ucc98\ub9ac\ud558\ub294 \ubaa8\ub378\uc758 \ub2a5\ub825\uc744 \ud5a5\uc0c1\uc2dc\ud0a4\uae30 \uc704\ud574 \ubcf4\ub2e4 \uc815\uad50\ud55c \uc815\uaddc\ud654 \ubc29\ubc95\uc744 \ud0d0\uc0c9\ud560 \uc218 \uc788\uc2b5\ub2c8\ub2e4.<\/p>\n<\/li>\n

            2. \n

              \uc2e0\uacbd\ub9dd \ud1b5\ud569<\/strong>: \uc131\ub2a5 \ubc0f \uc720\uc5f0\uc131 \ud5a5\uc0c1\uc744 \uc704\ud574 \uc2e0\uacbd\ub9dd \ubc0f \ub525 \ub7ec\ub2dd \uc544\ud0a4\ud14d\ucc98\ub97c LightGBM\uacfc \uac19\uc740 \uadf8\ub798\ub514\uc5b8\ud2b8 \ubd80\uc2a4\ud305 \ud504\ub808\uc784\uc6cc\ud06c\uc640 \ud1b5\ud569\ud558\ub824\ub294 \uc2dc\ub3c4\uac00 \uc788\uc744 \uc218 \uc788\uc2b5\ub2c8\ub2e4.<\/p>\n<\/li>\n

            3. \n

              AutoML \ud1b5\ud569<\/strong>: LightGBM\uc740 \uc790\ub3d9\ud654\ub41c \uae30\uacc4 \ud559\uc2b5(AutoML) \ud50c\ub7ab\ud3fc\uc5d0 \ud1b5\ud569\ub418\uc5b4 \ube44\uc804\ubb38\uac00\ub3c4 \ub2e4\uc591\ud55c \uc791\uc5c5\uc5d0 \uadf8 \ud798\uc744 \ud65c\uc6a9\ud560 \uc218 \uc788\uc2b5\ub2c8\ub2e4.<\/p>\n<\/li>\n

            4. \n

              \ubd84\uc0b0 \ucef4\ud4e8\ud305 \uc9c0\uc6d0<\/strong>: LightGBM\uc744 Apache Spark\uc640 \uac19\uc740 \ubd84\uc0b0 \ucef4\ud4e8\ud305 \ud504\ub808\uc784\uc6cc\ud06c\uc5d0\uc11c \uc2e4\ud589\ud560 \uc218 \uc788\ub3c4\ub85d \ub178\ub825\ud558\uba74 \ube45 \ub370\uc774\ud130 \uc2dc\ub098\ub9ac\uc624\uc758 \ud655\uc7a5\uc131\uc744 \ub354\uc6b1 \ud5a5\uc0c1\uc2dc\ud0ac \uc218 \uc788\uc2b5\ub2c8\ub2e4.<\/p>\n<\/li>\n<\/ol>\n

              \ud504\ub85d\uc2dc \uc11c\ubc84\ub97c \uc0ac\uc6a9\ud558\uac70\ub098 LightGBM\uacfc \uc5f0\uacb0\ud558\ub294 \ubc29\ubc95<\/h2>\n

              \ud504\ub85d\uc2dc \uc11c\ubc84\ub294 \ub2e4\uc591\ud55c \uc2dc\ub098\ub9ac\uc624\uc5d0\uc11c LightGBM\uc744 \uc0ac\uc6a9\ud560 \ub54c \uc911\uc694\ud55c \uc5ed\ud560\uc744 \ud560 \uc218 \uc788\uc2b5\ub2c8\ub2e4.<\/p>\n

                \n
              1. \n

                \ub370\uc774\ud130 \uc2a4\ud06c\ub798\ud551<\/strong>: \uae30\uacc4 \ud559\uc2b5 \uc791\uc5c5\uc744 \uc704\ud55c \ub370\uc774\ud130\ub97c \uc218\uc9d1\ud560 \ub54c \ud504\ub85d\uc2dc \uc11c\ubc84\ub97c \uc0ac\uc6a9\ud558\uc5ec IP \ucc28\ub2e8 \ub610\ub294 \uc18d\ub3c4 \uc81c\ud55c \ubb38\uc81c\ub97c \ubc29\uc9c0\ud558\uba74\uc11c \uc6f9\uc0ac\uc774\ud2b8\uc5d0\uc11c \uc815\ubcf4\ub97c \uc2a4\ud06c\ub7a9\ud560 \uc218 \uc788\uc2b5\ub2c8\ub2e4.<\/p>\n<\/li>\n

              2. \n

                \ub370\uc774\ud130 \ud504\ub77c\uc774\ubc84\uc2dc<\/strong>: \ud504\ub85d\uc2dc \uc11c\ubc84\ub294 \ud2b9\ud788 \ub370\uc774\ud130 \ubcf4\ud638\uac00 \uc911\uc694\ud55c \uc560\ud50c\ub9ac\ucf00\uc774\uc158\uc5d0\uc11c \ubaa8\ub378 \ud6c8\ub828 \uc911\uc5d0 \uc0ac\uc6a9\uc790\uc758 IP \uc8fc\uc18c\ub97c \uc775\uba85\ud654\ud558\uc5ec \ub370\uc774\ud130 \uac1c\uc778 \uc815\ubcf4 \ubcf4\ud638\ub97c \uac15\ud654\ud560 \uc218 \uc788\uc2b5\ub2c8\ub2e4.<\/p>\n<\/li>\n

              3. \n

                \ubd84\uc0b0 \uad50\uc721<\/strong>: \ubd84\uc0b0\ub41c \uae30\uacc4 \ud559\uc2b5 \uc124\uc815\uc758 \uacbd\uc6b0 \ud504\ub85d\uc2dc \uc11c\ubc84\ub97c \ud65c\uc6a9\ud558\uc5ec \ub178\ub4dc \uac04 \ud1b5\uc2e0\uc744 \uad00\ub9ac\ud558\uace0 \uc5ec\ub7ec \uc704\uce58\uc5d0\uc11c \ud611\uc5c5 \uad50\uc721\uc744 \ucd09\uc9c4\ud560 \uc218 \uc788\uc2b5\ub2c8\ub2e4.<\/p>\n<\/li>\n

              4. \n

                \ub85c\ub4dc \ubc38\ub7f0\uc2f1<\/strong>: \ud504\ub85d\uc2dc \uc11c\ubc84\ub294 \ub4e4\uc5b4\uc624\ub294 \uc694\uccad\uc744 \uc5ec\ub7ec LightGBM \uc778\uc2a4\ud134\uc2a4\uc5d0 \ubd84\uc0b0\ud558\uc5ec \uacc4\uc0b0 \ub9ac\uc18c\uc2a4 \uc0ac\uc6a9\uc744 \ucd5c\uc801\ud654\ud558\uace0 \uc804\ubc18\uc801\uc778 \uc131\ub2a5\uc744 \ud5a5\uc0c1\uc2dc\ud0ac \uc218 \uc788\uc2b5\ub2c8\ub2e4.<\/p>\n<\/li>\n<\/ol>\n

                \uad00\ub828\ub41c \ub9c1\ud06c\ub4e4<\/h2>\n

                LightGBM\uc5d0 \ub300\ud55c \uc790\uc138\ud55c \ub0b4\uc6a9\uc744 \ubcf4\ub824\uba74 \ub2e4\uc74c \ub9ac\uc18c\uc2a4\ub97c \uc0b4\ud3b4\ubcf4\uc138\uc694.<\/p>\n

                  \n
                1. \n

                  \uacf5\uc2dd LightGBM GitHub \ub9ac\ud3ec\uc9c0\ud1a0\ub9ac<\/a>: LightGBM\uc758 \uc18c\uc2a4 \ucf54\ub4dc, \ubb38\uc11c, \uc774\uc288 \ud2b8\ub798\ucee4\uc5d0 \uc561\uc138\uc2a4\ud558\uc138\uc694.<\/p>\n<\/li>\n

                2. \n

                  LightGBM\uc5d0 \uad00\ud55c Microsoft \uc5f0\uad6c \ub17c\ubb38<\/a>: LightGBM\uc744 \uc18c\uac1c\ud55c \uc6d0\ubcf8 \uc5f0\uad6c \ub17c\ubb38\uc744 \uc77d\uc5b4\ubcf4\uc138\uc694.<\/p>\n<\/li>\n

                3. \n

                  LightGBM \ubb38\uc11c<\/a>: \uc790\uc138\ud55c \uc0ac\uc6a9 \uc9c0\uce68, API \ucc38\uc870 \ubc0f \ud29c\ud1a0\ub9ac\uc5bc\uc740 \uacf5\uc2dd \ubb38\uc11c\ub97c \ucc38\uc870\ud558\uc138\uc694.<\/p>\n<\/li>\n

                4. \n

                  \uce90\uae00 \ub300\ud68c<\/a>: LightGBM\uc774 \ub110\ub9ac \uc0ac\uc6a9\ub418\ub294 Kaggle \ub300\ud68c\ub97c \uc0b4\ud3b4\ubcf4\uace0 \uc608\uc81c \ub178\ud2b8\ubd81 \ubc0f \ucee4\ub110\uc744 \ud1b5\ud574 \uc54c\uc544\ubcf4\uc138\uc694.<\/p>\n<\/li>\n<\/ol>\n

                  LightGBM\uc758 \uae30\ub2a5\uc744 \ud65c\uc6a9\ud558\uace0 \uadf8 \ubbf8\ubb18\ud55c \ucc28\uc774\ub97c \uc774\ud574\ud568\uc73c\ub85c\uc368 \ub370\uc774\ud130 \uacfc\ud559\uc790\uc640 \uc5f0\uad6c\uc790\ub294 \uae30\uacc4 \ud559\uc2b5 \ubaa8\ub378\uc744 \ud5a5\uc0c1\ud558\uace0 \ubcf5\uc7a1\ud55c \uc2e4\uc81c \ubb38\uc81c\ub97c \ud574\uacb0\ud558\ub294 \ub370 \uc788\uc5b4 \uacbd\uc7c1 \uc6b0\uc704\ub97c \ud655\ubcf4\ud560 \uc218 \uc788\uc2b5\ub2c8\ub2e4. \ub300\uaddc\ubaa8 \ub370\uc774\ud130 \ubd84\uc11d, \uc815\ud655\ud55c \uc608\uce21, \uac1c\uc778\ud654\ub41c \ucd94\ucc9c \ub4f1 LightGBM\uc740 \ud0c1\uc6d4\ud55c \uc18d\ub3c4\uc640 \ud6a8\uc728\uc131\uc73c\ub85c AI \ucee4\ubba4\ub2c8\ud2f0\uc5d0 \uc9c0\uc18d\uc801\uc73c\ub85c \ud798\uc744 \uc2e4\uc5b4\uc8fc\uace0 \uc788\uc2b5\ub2c8\ub2e4.<\/p>","protected":false},"featured_media":468775,"menu_order":0,"template":"","meta":{"_acf_changed":false,"content-type":"","inline_featured_image":false,"footnotes":""},"class_list":["post-477828","wiki","type-wiki","status-publish","has-post-thumbnail","hentry"],"acf":{"faq_title":"Frequently Asked Questions about LightGBM: Boosting Performance with Speed and Efficiency<\/mark>","faq_items":[{"question":"What is LightGBM?","answer":"

                  LightGBM is a powerful and efficient open-source machine learning library designed for gradient boosting. It is developed by Microsoft and is widely used for handling large-scale datasets with high accuracy.<\/p>"},{"question":"How did LightGBM originate?","answer":"

                  LightGBM was introduced in 2017 by Microsoft researchers in a paper titled \"LightGBM: A Highly Efficient Gradient Boosting Decision Tree.\" The paper presented LightGBM as a novel method for boosting efficiency in gradient boosting algorithms.<\/p>"},{"question":"How does LightGBM work?","answer":"

                  LightGBM operates on the gradient boosting technique with a leaf-wise tree growth strategy. It selects the leaf node with the maximum loss reduction during each tree expansion, resulting in a more accurate model with fewer leaves. The library optimizes memory usage through techniques like Gradient-based One-Side Sampling (GOSS) and Exclusive Feature Bundling (EFB).<\/p>"},{"question":"What are the key features of LightGBM?","answer":"

                  LightGBM boasts high speed, memory efficiency, scalability, and flexibility. Its leaf-wise tree growth strategy enhances predictive accuracy, and it supports various machine learning tasks, such as regression, classification, ranking, and recommendation systems.<\/p>"},{"question":"What types of LightGBM are there?","answer":"

                  LightGBM offers two main types: Gradient Boosting Machine (GBM) and Dart. GBM uses leaf-wise tree growth, while Dart includes dropout-based regularization to prevent overfitting.<\/p>"},{"question":"How can LightGBM be used?","answer":"

                  LightGBM is versatile and can be used for classification, regression, ranking, and recommendation systems. It is effective in handling large datasets and provides accurate predictions.<\/p>"},{"question":"What are the challenges in using LightGBM?","answer":"

                  Users may face challenges with imbalanced datasets, overfitting, hyperparameter tuning, and data preprocessing. However, solutions like class weights, cross-validation, and proper data handling can help mitigate these issues.<\/p>"},{"question":"How does LightGBM compare to other gradient boosting libraries?","answer":"

                  In comparison to XGBoost and CatBoost, LightGBM stands out with its faster speed and efficient memory usage. It excels in handling large datasets and offers similar performance to CatBoost.<\/p>"},{"question":"What does the future hold for LightGBM?","answer":"

                  The future of LightGBM may involve enhanced regularization techniques, integration with neural networks, AutoML support, and distributed computing capabilities to further improve its performance.<\/p>"},{"question":"How can proxy servers be associated with LightGBM?","answer":"

                  Proxy servers can be beneficial in data scraping, data privacy, distributed training, and load balancing when using LightGBM for machine learning tasks.<\/p>

                  For more detailed information, please refer to the article above.<\/p>"}]},"_links":{"self":[{"href":"https:\/\/oneproxy.pro\/kr\/wp-json\/wp\/v2\/wiki\/477828","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/oneproxy.pro\/kr\/wp-json\/wp\/v2\/wiki"}],"about":[{"href":"https:\/\/oneproxy.pro\/kr\/wp-json\/wp\/v2\/types\/wiki"}],"version-history":[{"count":0,"href":"https:\/\/oneproxy.pro\/kr\/wp-json\/wp\/v2\/wiki\/477828\/revisions"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/oneproxy.pro\/kr\/wp-json\/wp\/v2\/media\/468775"}],"wp:attachment":[{"href":"https:\/\/oneproxy.pro\/kr\/wp-json\/wp\/v2\/media?parent=477828"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}