XGBoost

["XGBoost(eXtreme Gradient Boosting)\ub294 C++, Java, Python, R, Julia, Perl, Scala\ub97c \uc704\ud55c \uc815\uaddc\ud654 \uadf8\ub798\ub514\uc5b8\ud2b8 \ubd80\uc2a4\ud305 \ud504\ub808\uc784\uc6cc\ud06c\ub97c \uc81c\uacf5\ud558\ub294 \uc624\ud508\uc18c\uc2a4 \uc18c\ud504\ud2b8\uc6e8\uc5b4 \ub77c\uc774\ube0c\ub7ec\ub9ac\uc774\ub2e4. Linux, Microsoft Windows, macOS\uc5d0\uc11c \uc791\ub3d9\ud55c\ub2e4. \uc774 \ud504\ub85c\uc81d\ud2b8\ub294 \"\ud655\uc7a5 \uac00\ub2a5\ud558\uace0, \ud734\ub300 \uac00\ub2a5\ud558\uba70, \ubd84\uc0b0\ub41c \uadf8\ub798\ub514\uc5b8\ud2b8 \ubd80\uc2a4\ud305(GBM, GBRT, GBDT) \ub77c\uc774\ube0c\ub7ec\ub9ac\"\ub97c \uc81c\uacf5\ud558\ub294 \uac83\uc744 \ubaa9\ud45c\ub85c \ud558\uba70, \ub2e8\uc77c \uba38\uc2e0\ubfd0\ub9cc \uc544\ub2c8\ub77c Apache Hadoop, Apache Spark, Apache Flink, Dask\uc640 \uac19\uc740 \ubd84\uc0b0 \ucc98\ub9ac \ud504\ub808\uc784\uc6cc\ud06c\uc5d0\uc11c\ub3c4 \uc2e4\ud589\ub41c\ub2e4. XGBoost\ub294 2010\ub144\ub300 \uc911\ubc18\uc5d0 \uba38\uc2e0\ub7ec\ub2dd \ub300\ud68c, \ud2b9\ud788 Kaggle\uacfc \uac19\uc740 \ud50c\ub7ab\ud3fc\uc5d0\uc11c \ub9ce\uc740 \uc6b0\uc2b9 \ud300\uc774 \uc120\ud0dd\ud55c \uc54c\uace0\ub9ac\uc998\uc73c\ub85c \ud070 \uc778\uae30\ub97c \uc5bb\uc5c8\ub2e4.\n\n\uc774 \ub77c\uc774\ube0c\ub7ec\ub9ac\ub294 \uc77c\ubc18\uc801\uc73c\ub85c \uc758\uc0ac \uacb0\uc815 \ud2b8\ub9ac\uc640 \uac19\uc740 \uc5ec\ub7ec \uc57d\ud55c \uc608\uce21 \ubaa8\ub378\uc744 \ud558\ub098\uc758 \uac15\ub825\ud55c \ubaa8\ub378\ub85c \uacb0\ud569\ud558\ub294 \uc559\uc0c1\ube14 \uae30\ubc95\uc778 \uadf8\ub798\ub514\uc5b8\ud2b8 \ubd80\uc2a4\ud305\uc758 \uc6d0\ub9ac\ub97c \uae30\ubc18\uc73c\ub85c \ud55c\ub2e4. XGBoost\ub294 \uc18d\ub3c4, \ud655\uc7a5\uc131, \uc815\uaddc\ud654 \uce21\uba74\uc758 \ucd5c\uc801\ud654\ub97c \ud1b5\ud574 \ucc28\ubcc4\ud654\ub418\uba70, \uae08\uc735\uc5d0\uc11c \uc758\ub8cc\uc5d0 \uc774\ub974\uae30\uae4c\uc9c0 \ub2e4\uc591\ud55c \ubd84\uc57c\uc758 \ubd84\ub958, \ud68c\uadc0, \uc21c\uc704 \uc791\uc5c5\uc5d0 \uc720\uc6a9\ud55c \ub3c4\uad6c\uac00 \ub418\uc5c8\ub2e4.\n\n## \uc5ed\uc0ac\n\nXGBoost\ub294 Tianqi Chen\uc774 \ud1a0\ub860\ud1a0 \ub300\ud559\uad50\uc758 DMLC(Distributed (Deep) Machine Learning Community) \uadf8\ub8f9\uc758 \uc77c\ubd80\ub85c \uac1c\ubc1c\ud55c \uc5f0\uad6c \ud504\ub85c\uc81d\ud2b8\uc5d0\uc11c \uc2dc\uc791\ub418\uc5c8\ub2e4(\ucd08\uae30 \uc791\uc5c5\uc740 \uc6cc\uc2f1\ud134 \ub300\ud559\uad50\uc5d0\uc11c \uc774\ub8e8\uc5b4\uc9d0). libsvm \uad6c\uc131 \ud30c\uc77c\uc744 \uc0ac\uc6a9\ud558\uc5ec \uad6c\uc131\ub41c \ud130\ubbf8\ub110 \uc560\ud50c\ub9ac\ucf00\uc774\uc158\uc73c\ub85c \uc2dc\uc791\ub418\uc5c8\ub2e4. \uc774 \ud504\ub85c\uc81d\ud2b8\ub294 CERN\uacfc \ub2e4\ub978 \uae30\uad00\ub4e4\uc774 \uc785\uc790 \uc0ac\uac74\uc744 \ubd84\ub958\ud558\uae30 \uc704\ud574 \uc870\uc9c1\ud55c \ub300\ud68c\uc778 Higgs Machine Learning Challenge\uc758 \uc6b0\uc2b9 \uc194\ub8e8\uc158\uc5d0 \uc0ac\uc6a9\ub41c \ud6c4 \uba38\uc2e0\ub7ec\ub2dd \ub300\ud68c \ubd84\uc57c\uc5d0\uc11c \uc778\uc815\uc744 \ubc1b\uc558\ub2e4. \uc774\ub7ec\ud55c \uc131\uacf5\uc73c\ub85c Python \ubc0f R \ud328\ud0a4\uc9c0\uac00 \ube60\ub974\uac8c \uac1c\ubc1c\ub418\uc5c8\uace0, \uc774\ud6c4 Java, Scala, Julia, Perl \ubc0f \uae30\ud0c0 \uc5b8\uc5b4\uc6a9 \uad6c\ud604\uc774 \uc774\uc5b4\uc838 \uc0ac\uc6a9\uc790 \uae30\ubc18\uc774 \ub113\uc5b4\uc9c0\uace0 Kaggle \ucee4\ubba4\ub2c8\ud2f0\uc5d0\uc11c \uc778\uae30\ub97c \uc5bb\ub294 \ub370 \uae30\uc5ec\ud588\ub2e4.\n\nXGBoost\ub294 \ucc44\ud0dd\uc744 \uc6a9\uc774\ud558\uac8c \ud558\uae30 \uc704\ud574 \uace7 \ub2e4\ub978 \ud328\ud0a4\uc9c0\uc640 \ud1b5\ud569\ub418\uc5c8\ub2e4. Python \uc0ac\uc6a9\uc790\ub97c \uc704\ud55c scikit-learn \ubc0f R \uc0ac\uc6a9\uc790\ub97c \uc704\ud55c caret \ud328\ud0a4\uc9c0\uc640 \ud568\uaed8 \uc0ac\uc6a9\ud560 \uc218 \uc788\uac8c \ub418\uc5c8\ub2e4. Apache Spark, Apache Hadoop, Apache Flink\uc640 \uac19\uc740 \ub370\uc774\ud130 \ud750\ub984 \ud504\ub808\uc784\uc6cc\ud06c\uc640\uc758 \ud1b5\ud569\uc740 \ucd94\uc0c1\ud654\ub41c Rabit \ubc0f XGBoost4J \uc778\ud130\ud398\uc774\uc2a4\ub97c \ud1b5\ud574 \uc774\ub8e8\uc5b4\uc84c\ub2e4. \ub610\ud55c XGBoost\ub294 FPGA\uc6a9 OpenCL\uc5d0\uc11c\ub3c4 \uc0ac\uc6a9\ud560 \uc218 \uc788\ub2e4. \ud6a8\uc728\uc801\uc774\uace0 \ud655\uc7a5 \uac00\ub2a5\ud55c \uad6c\ud604\uc740 Tianqi Chen\uacfc Carlos Guestrin\uc774 \ubc1c\ud45c\ud588\uc73c\uba70, \uc54c\uace0\ub9ac\uc998 \ubc0f \uc2dc\uc2a4\ud15c \ucd5c\uc801\ud654\uc5d0 \ub300\ud574 \uc790\uc138\ud788 \uc124\uba85\ud588\ub2e4.\n\nXGBoost\ub294 \ub2e8\uc77c \uc758\uc0ac \uacb0\uc815 \ud2b8\ub9ac\ubcf4\ub2e4 \ub354 \ub192\uc740 \uc815\ud655\ub3c4\ub97c \ub2ec\uc131\ud558\ub294 \uacbd\uc6b0\uac00 \ub9ce\uc9c0\ub9cc, \uc758\uc0ac \uacb0\uc815 \ud2b8\ub9ac\uc758 \uace0\uc720\ud55c \ud574\uc11d \uac00\ub2a5\uc131\uc744 \ud76c\uc0dd\ud55c\ub2e4. \ub2e8\uc77c \ud2b8\ub9ac\uac00 \uacb0\uc815\uc744 \ub0b4\ub9ac\uae30 \uc704\ud574 \ub530\ub974\ub294 \uacbd\ub85c\ub97c \ucd94\uc801\ud558\ub294 \uac83\uc740 \uac04\ub2e8\ud558\uace0 \uc790\uba85\ud558\uc9c0\ub9cc, \uc218\ubc31 \ub610\ub294 \uc218\ucc9c \uac1c\uc758 \ud2b8\ub9ac\uc758 \uacbd\ub85c\ub97c \ucd94\uc801\ud558\ub294 \uac83\uc740 \ud6e8\uc52c \ub354 \uc5b4\ub824\uc6cc \ubaa8\ub378 \uc124\uba85\uc774 \ub354 \ubcf5\uc7a1\ud574\uc9c4\ub2e4.\n\n## \ud2b9\uc9d5\n\nXGBoost\ub294 \ub2e4\ub978 \uadf8\ub798\ub514\uc5b8\ud2b8 \ubd80\uc2a4\ud305 \uc54c\uace0\ub9ac\uc998\uacfc \ucc28\ubcc4\ud654\ub418\ub294 \uba87 \uac00\uc9c0 \ub450\ub4dc\ub7ec\uc9c4 \ud2b9\uc9d5\uc744 \ud3ec\ud568\ud55c\ub2e4:\n\n- \uacfc\uc801\ud569\uc744 \uc904\uc774\uae30 \uc704\ud574 \uc815\uaddc\ud654\ub97c \uc801\uc6a9\ud558\ub294 \ud2b8\ub9ac\uc758 \uc601\ub9ac\ud55c \ud398\ub110\ud2f0 \ubd80\uc5ec.\n- \uac01 \ud2b8\ub9ac\uc758 \uae30\uc5ec\ub3c4\ub97c \uc870\uc815\ud558\ub294 \ub9ac\ud504 \ub178\ub4dc\uc758 \ube44\ub840 \ucd95\uc18c.\n- \ucd5c\uc801\ud654\uc5d0 2\ucc28 \ub3c4\ud568\uc218\ub97c \uc0ac\uc6a9\ud558\ub294 Newton Boosting.\n- \ud2b8\ub9ac \uac04 \uc0c1\uad00 \uad00\uacc4\ub97c \uc904\uc774\uae30 \uc704\ud55c \ucd94\uac00 \ubb34\uc791\uc704\ud654 \ub9e4\uac1c\ubcc0\uc218.\n- \ub2e8\uc77c, \ubd84\uc0b0 \uc2dc\uc2a4\ud15c \ubc0f \ub300\uaddc\ubaa8 \ub370\uc774\ud130 \uc138\ud2b8\ub97c \uc704\ud55c out-of-core \uacc4\uc0b0 \uad6c\ud604.\n- \ud6c8\ub828 \uc911 \uc790\ub3d9 \ud2b9\uc9d5 \uc120\ud0dd.\n- \ub300\uaddc\ubaa8 \ub370\uc774\ud130\uc758 \ud6a8\uc728\uc801\uc778 \uacc4\uc0b0\uc744 \uc704\ud55c \uc774\ub860\uc801\uc73c\ub85c \uc815\ub2f9\ud654\ub41c \uac00\uc911 \ubd84\uc704\uc218 \uc2a4\ucf00\uce58.\n- \ud76c\uc18c\uc131 \uc778\uc2dd\uc774 \uc788\ub294 \ubcd1\ub82c \ud2b8\ub9ac \uad6c\uc870 \ubd80\uc2a4\ud305\uc73c\ub85c \ub204\ub77d\ub41c \uac12\uc744 \ud6a8\uacfc\uc801\uc73c\ub85c \ucc98\ub9ac.\n- \uc758\uc0ac \uacb0\uc815 \ud2b8\ub9ac \ud6c8\ub828\uc744 \uc704\ud55c \ud6a8\uc728\uc801\uc778 \uce90\uc2dc \uac00\ub2a5 \ube14\ub85d \uad6c\uc870\ub85c \uba54\ubaa8\ub9ac \uc811\uadfc \ud328\ud134 \uac1c\uc120.\n\n\uc774\ub7ec\ud55c \ud2b9\uc9d5\uc740 \ub2e4\uc591\ud55c \ud658\uacbd\uc5d0\uc11c XGBoost\uc758 \ub192\uc740 \uc131\ub2a5\uacfc \uacac\uace0\uc131\uc5d0 \ub300\ud55c \ud3c9\ud310\uc5d0 \uae30\uc5ec\ud55c\ub2e4.\n\n## \uc54c\uace0\ub9ac\uc998\n\nXGBoost\ub294 \ud45c\uc900 \uadf8\ub798\ub514\uc5b8\ud2b8 \ubd80\uc2a4\ud305\uc774 \ud568\uc218 \uacf5\uac04\uc5d0\uc11c \uacbd\uc0ac \ud558\uac15\ubc95\uc73c\ub85c \uc791\ub3d9\ud558\ub294 \uac83\uacfc \ub2ec\ub9ac \ud568\uc218 \uacf5\uac04\uc5d0\uc11c Newton-Raphson\uc73c\ub85c \uc791\ub3d9\ud55c\ub2e4. \uc190\uc2e4 \ud568\uc218\uc5d0 2\ucc28 Taylor \uadfc\uc0ac\uac00 \uc0ac\uc6a9\ub418\uc5b4 Newton-Raphson \ubc29\ubc95\uacfc\uc758 \uc5f0\uacb0\uc744 \uc124\uc815\ud55c\ub2e4. \uc774 \uc811\uadfc \ubc29\uc2dd\uc744 \ud1b5\ud574 \uc54c\uace0\ub9ac\uc998\uc740 \uace1\ub960 \uc815\ubcf4\ub97c \ud3ec\ucc29\ud560 \uc218 \uc788\uc5b4 \ub354 \ube60\ub978 \uc218\ub834\uacfc \uc885\uc885 \ub354 \ub098\uc740 \uc815\ud655\ub3c4\ub97c \uc5bb\uc744 \uc218 \uc788\ub2e4.\n\n\uc77c\ubc18\uc801\uc778 \ube44\uc815\uaddc\ud654 XGBoost \uc54c\uace0\ub9ac\uc998\uc740 \uc190\uc2e4 \ud568\uc218\ub97c \ucd5c\uc18c\ud654\ud558\uae30 \uc704\ud574 \ud2b8\ub9ac\ub97c \ubc18\ubcf5\uc801\uc73c\ub85c \ucd94\uac00\ud55c\ub2e4. \uac01 \ub2e8\uacc4\uc5d0\uc11c \uc54c\uace0\ub9ac\uc998\uc740 \ud604\uc7ac \uc608\uce21\uc5d0 \ub300\ud55c \uc190\uc2e4\uc758 \uadf8\ub798\ub514\uc5b8\ud2b8\uc640 \ud5e4\uc2dc\uc548\uc744 \uacc4\uc0b0\ud55c \ub2e4\uc74c \uc774\ub7ec\ud55c \uac12\uc5d0 \ud2b8\ub9ac\ub97c \ud53c\ud305\ud55c\ub2e4. \ud2b8\ub9ac \uad6c\uc870\ub294 \uc190\uc2e4 \uac10\uc18c\ub97c \ucd5c\ub300\ud654\ud558\ub294 \ubd84\ud560 \ud6c4\ubcf4\ub97c \ud3c9\uac00\ud558\uc5ec \ud559\uc2b5\ub418\uba70, \uc815\uaddc\ud654 \ud56d\uc774 \ubcf5\uc7a1\uc131\uc744 \uc81c\uc5b4\ud55c\ub2e4.\n\n\ud76c\uc18c\uc131\uc740 \ub204\ub77d\ub41c \uac12\uc774 \ud6c8\ub828 \ub370\uc774\ud130\ub97c \uae30\ubc18\uc73c\ub85c \ucd5c\uc801\uc758 \ubd84\uae30\ub85c \ub77c\uc6b0\ud305\ub418\ub294 \uae30\ubcf8 \ubc29\ud5a5 \uba54\ucee4\ub2c8\uc998\uc744 \ud1b5\ud574 \ucc98\ub9ac\ub41c\ub2e4. \ubcd1\ub82c \ud2b8\ub9ac \ubd80\uc2a4\ud305\uc740 \ud6a8\uc728\uc801\uc778 \uc5f4 \ubc29\ud5a5 \uc811\uadfc\uc744 \uac00\ub2a5\ud558\uac8c \ud558\ub294 \ube14\ub85d \uad6c\uc870\ub97c \uc0ac\uc6a9\ud558\uc5ec \uad6c\ud604\ub418\uba70, out-of-core \uacc4\uc0b0 \ubc0f \ubd84\uc0b0 \ud6c8\ub828\uc744 \uc9c0\uc6d0\ud55c\ub2e4.\n\n## \ub9e4\uac1c\ubcc0\uc218\n\nXGBoost\ub294 \ub3d9\uc791\uacfc \uc131\ub2a5\uc5d0 \uc601\ud5a5\uc744 \ubbf8\uce58\ub294 \uc218\ub9ce\uc740 \ub9e4\uac1c\ubcc0\uc218\ub97c \ub178\ucd9c\ud55c\ub2e4. \uc8fc\uc694 \ub9e4\uac1c\ubcc0\uc218\ub294 \ub2e4\uc74c\uacfc \uac19\ub2e4:\n\n- \ud559\uc2b5\ub960(\"step size\" \ub610\ub294 \"shrinkage\"\ub77c\uace0\ub3c4 \ud568): 0\uacfc 1 \uc0ac\uc774\uc758 \uc22b\uc790\ub85c, \uae30\ubcf8\uac12\uc740 0.3\uc774\uba70 \uc54c\uace0\ub9ac\uc998\uc774 \uac01 \ubc18\ubcf5\uc5d0\uc11c \ud559\uc2b5\ud558\ub294 \uc815\ub3c4\ub97c \uacb0\uc815\ud55c\ub2e4. \uac12\uc774 \ub0ae\uc744\uc218\ub85d \ub354 \ub9ce\uc740 \ud2b8\ub9ac\uac00 \ud544\uc694\ud558\uc9c0\ub9cc \uc77c\ubc18\ud654\ub97c \uac1c\uc120\ud560 \uc218 \uc788\ub2e4.\n- n_estimators: \uc559\uc0c1\ube14\uc5d0 \uad6c\ucd95\ud560 \ud2b8\ub9ac \uc218\ub97c \uc124\uc815\ud55c\ub2e4. \ud2b8\ub9ac\uac00 \ub9ce\uc744\uc218\ub85d \ubaa8\ub378 \ubcf5\uc7a1\uc131\uc774 \uc99d\uac00\ud558\uc9c0\ub9cc \ub108\ubb34 \ub9ce\uc73c\uba74 \uacfc\uc801\ud569\uc73c\ub85c \uc774\uc5b4\uc9c8 \uc218 \uc788\ub2e4.\n- Gamma(Lagrange multiplier \ub610\ub294 \ucd5c\uc18c \uc190\uc2e4 \uac10\uc18c \ub9e4\uac1c\ubcc0\uc218\ub77c\uace0\ub3c4 \ud568): \ub9ac\ud504 \ub178\ub4dc\uc5d0\uc11c \ucd94\uac00 \ubd84\ud560\uc744 \uc218\ud589\ud558\ub294 \ub370 \ud544\uc694\ud55c \ucd5c\uc18c \uc190\uc2e4 \uac10\uc18c\ub7c9\uc744 \uc81c\uc5b4\ud55c\ub2e4. \uae30\ubcf8\uac12\uc740 0\uc774\ub2e4.\n- max_depth: \ud6c8\ub828 \uc911 \uac01 \ud2b8\ub9ac\uac00 \uc5bc\ub9c8\ub098 \uae4a\uc774 \uc131\uc7a5\ud560 \uc218 \uc788\ub294\uc9c0\ub97c \ub098\ud0c0\ub0b4\uba70 \uae30\ubcf8\uac12\uc740 6\uc774\ub2e4. \ub354 \uae4a\uc740 \ud2b8\ub9ac\ub294 \ub354 \ubcf5\uc7a1\ud55c \ud328\ud134\uc744 \ud3ec\ucc29\ud558\uc9c0\ub9cc \uacfc\uc801\ud569 \uc704\ud5d8\uc774 \uc788\ub2e4.\n\n\uae30\ud0c0 \ub9e4\uac1c\ubcc0\uc218\ub85c\ub294 subsample, colsample_bytree, reg_alpha, reg_lambda\uac00 \uc788\uc73c\uba70, \uc815\uaddc\ud654 \ubc0f \uc0d8\ud50c\ub9c1\uc5d0 \ub300\ud55c \ucd94\uac00 \uc81c\uc5b4\ub97c \uc81c\uacf5\ud55c\ub2e4.\n\n## \uc751\uc6a9 \ubc0f \uc601\ud5a5\n\nXGBoost\ub294 \uc0b0\uc5c5 \ubc0f \ud559\uacc4\uc5d0\uc11c \ub110\ub9ac \ucc44\ud0dd\ub418\uc5c8\ub2e4. \uae08\uc735\uc5d0\uc11c\ub294 \uc2e0\uc6a9 \ud3c9\uac00, \uc0ac\uae30 \ud0d0\uc9c0, \uc704\ud5d8 \ubaa8\ub378\ub9c1\uc5d0 \uc0ac\uc6a9\ub41c\ub2e4. \uc758\ub8cc\uc5d0\uc11c\ub294 \uc9c8\ubcd1 \uc608\uce21 \ubc0f \ud658\uc790 \uacb0\uacfc \ubd84\uc11d\uc744 \uc9c0\uc6d0\ud55c\ub2e4. \uc804\uc790 \uc0c1\uac70\ub798\uc5d0\uc11c\ub294 \ucd94\ucc9c \uc2dc\uc2a4\ud15c\uacfc \uace0\uac1d \uc774\ud0c8 \uc608\uce21\uc744 \uac15\ud654\ud55c\ub2e4. Kaggle\uacfc \uac19\uc740 \ub300\ud68c\uc5d0\uc11c\uc758 \uc131\ub2a5\uc740 \ud45c \ud615\uc2dd \ub370\uc774\ud130 \ubb38\uc81c\uc758 \ubca4\uce58\ub9c8\ud06c\uac00 \ub418\uc5c8\ub2e4.\n\n\uc774 \ub77c\uc774\ube0c\ub7ec\ub9ac\ub294 Machine learning \ud504\ub808\uc784\uc6cc\ud06c\uc640\uc758 \ud1b5\ud569 \ubc0f \ubd84\uc0b0 \ucef4\ud4e8\ud305 \uc9c0\uc6d0\uc744 \ud1b5\ud574 \ub300\uaddc\ubaa8 \uc751\uc6a9 \ud504\ub85c\uadf8\ub7a8\uc5d0\uc11c \uc0ac\uc6a9\ud560 \uc218 \uc788\uac8c \ub418\uc5c8\ub2e4. Amazon Web Services \ubc0f Google Cloud\uc640 \uac19\uc740 \ud50c\ub7ab\ud3fc\uc758 \uad00\ub9ac\ud615 \uba38\uc2e0\ub7ec\ub2dd \uc11c\ube44\uc2a4\uc5d0 \ud1b5\ud569\ub418\uc5c8\ub2e4.\n\n## \uc218\uc0c1 \ubc0f \uc778\uc815\n\nXGBoost\ub294 2016\ub144 John Chambers Award, 2016\ub144 High Energy Physics meets Machine Learning award(HEP meets ML), KDD 2026\uc5d0\uc11c \"Test of Time Award\"\ub97c \ud3ec\ud568\ud55c \uc5ec\ub7ec \uc0c1\uc744 \ubc1b\uc558\ub2e4. \uc774\ub7ec\ud55c \uc778\uc815\uc740 \uba38\uc2e0\ub7ec\ub2dd\uc758 \uc751\uc6a9 \ubc0f \uc774\ub860\uc801 \uce21\uba74 \ubaa8\ub450\uc5d0 \ub300\ud55c \uae30\uc5ec\ub97c \uac15\uc870\ud55c\ub2e4.\n\n## \uac19\uc774 \ubcf4\uae30\n\n- \uba38\uc2e0\ub7ec\ub2dd \uc18c\ud504\ud2b8\uc6e8\uc5b4 \ube44\uad50\n- TabPFN\n- LightGBM\n- CatBoost\n\n## \ucc38\uace0 \ubb38\ud5cc\n\n- Chen, T., & Guestrin, C. (2016). XGBoost: A Scalable Tree Boosting System. Proceedings of the 22nd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining.\n- \uacf5\uc2dd \uc800\uc7a5\uc18c\uc5d0\uc11c \uc81c\uacf5\ub418\ub294 \ud504\ub85c\uc81d\ud2b8 \ubb38\uc11c \ubc0f \uc18c\uc2a4 \ucf54\ub4dc.", true]