\uc778\uacf5\uc9c0\ub2a5(AI)\uacfc \uba38\uc2e0\ub7ec\ub2dd(ML)\uc774 \ubc1c\uc804\ud568\uc5d0 \ub530\ub77c \ud559\uc2b5 \uc2dc\uc2a4\ud15c\uc758 \uc801\uc751\uc131\uacfc \uc720\uc5f0\uc131\uc5d0 \ub300\ud55c \uad00\uc2ec\uc774 \uc99d\uac00\ud558\uace0 \uc788\uc2b5\ub2c8\ub2e4. \uc9c0\uc18d \ud559\uc2b5(Continual Learning)\uacfc \ud3c9\uc0dd \ud559\uc2b5(Lifelong Learning)\uc740 \uc774\ub7ec\ud55c \ubc1c\uc804\uc744 \uac00\ub2a5\ud558\uac8c \ud558\ub294 \uc911\uc694\ud55c \uac1c\ub150\uc73c\ub85c \uc790\ub9ac \uc7a1\uace0 \uc788\uc2b5\ub2c8\ub2e4. \uc774 \uae00\uc5d0\uc11c\ub294 \ub450 \uac1c\ub150\uc758 \uc815\uc758\uc640 \ud544\uc694\uc131, \uadf8\ub9ac\uace0 \ub3c4\uc804 \uacfc\uc81c\ub97c \uc790\uc138\ud788 \uc0b4\ud3b4\ubcf4\uaca0\uc2b5\ub2c8\ub2e4.<\/p>\n
\uc9c0\uc18d \ud559\uc2b5\uc740 \uc2dc\uc2a4\ud15c\uc774 \uc0c8\ub85c\uc6b4 \uc815\ubcf4\ub97c \uc9c0\uc18d\uc801\uc73c\ub85c \ud559\uc2b5\ud558\uba74\uc11c\ub3c4 \uc774\uc804\uc5d0 \ud559\uc2b5\ud55c \uc815\ubcf4\ub97c \uc78a\uc9c0 \uc54a\ub294 \ub2a5\ub825\uc744 \uac00\ub9ac\ud0b5\ub2c8\ub2e4. \uc774\ub294 \ud2b9\ud788 \uc5b4\ub5a4 \uc0c1\ud669\uc5d0\uc11c\ub294 \uc774\uc804\uc758 \uc9c0\uc2dd\uc774 \uc5ec\uc804\ud788 \uc720\uc6a9\ud560 \uc218 \uc788\uae30 \ub54c\ubb38\uc5d0 \uc911\uc694\ud569\ub2c8\ub2e4. \uc608\ub97c \ub4e4\uc5b4, \uc790\uc728\uc8fc\ud589\ucc28\ub294 \ub2e4\uc591\ud55c \ud658\uacbd\uc5d0\uc11c \uc6b4\uc804 \uae30\uc220\uc744 \uc2b5\ub4dd\ud574\uc57c \ud558\uba70, \ud55c \ubc88 \uc775\ud78c \uae30\uc220\uc744 \uc78a\uc73c\uba74 \uc548 \ub429\ub2c8\ub2e4.<\/p>\n
\ud3c9\uc0dd \ud559\uc2b5\uc740 \uac1c\uc778\uc774\ub098 \uc2dc\uc2a4\ud15c\uc774 \uc0dd\uc560 \uc8fc\uae30 \uc804\uccb4\uc5d0 \uac78\uccd0 \uc9c0\uc18d\uc801\uc73c\ub85c \ud559\uc2b5\ud558\uace0 \uc801\uc751\ud558\ub294 \uacfc\uc815\uc744 \uac00\ub9ac\ud0b5\ub2c8\ub2e4. \uc774\ub294 \ub2e8\uc9c0 \ud2b9\uc815 \uc791\uc5c5\uc774\ub098 \ubb38\uc81c \ud574\uacb0\uc744 \ub118\uc5b4\uc11c, \ub2e4\uc591\ud55c \ub3c4\uba54\uc778\uacfc \uc0c1\ud669\uc5d0\uc11c\ub3c4 \uc801\uc6a9\ub420 \uc218 \uc788\ub294 \uc77c\ubc18\uc801\uc778 \ud559\uc2b5 \ub2a5\ub825\uc744 \ud3ec\ud568\ud569\ub2c8\ub2e4. \ud3c9\uc0dd \ud559\uc2b5 \uc2dc\uc2a4\ud15c\uc740 \uc9c0\uc18d\uc801\uc73c\ub85c \ud559\uc2b5\ud558\uc5ec \uc0c8\ub85c\uc6b4 \uc9c0\uc2dd\uc744 \uc313\uace0, \uc774\ub97c \uae30\ubc18\uc73c\ub85c \ubc1c\uc804\ud560 \uc218 \uc788\ub294 \ub2a5\ub825\uc744 \uac00\uc9c0\uace0 \uc788\uc2b5\ub2c8\ub2e4.<\/p>\n
\uc9c0\uc18d \ud559\uc2b5\uc744 \uc704\ud55c \uc5ec\ub7ec \uc811\uadfc\ubc95\ub4e4\uc774 \uc81c\uc548\ub418\uace0 \uc788\uc2b5\ub2c8\ub2e4. \uc608\ub97c \ub4e4\uba74, \ub2e4\uc74c\uacfc \uac19\uc740 \ubc29\ubc95\ub4e4\uc774 \uc788\uc2b5\ub2c8\ub2e4:<\/p>\n
\ub2e4\uc74c\uc740 \uac04\ub2e8\ud55c \uc9c0\uc18d \ud559\uc2b5 \uae30\ubc95\uc744 \uad6c\ud604\ud55c \uc608\uc81c\uc785\ub2c8\ub2e4. \uc774 \uc608\uc81c\uc5d0\uc11c\ub294 \ud30c\ub77c\ubbf8\ud130 \ud504\ub9ac\uc9d5 \uae30\ubc95\uc744 \uc0ac\uc6a9\ud558\uc5ec \uc9c0\uc18d \ud559\uc2b5\uc744 \uc2dc\ub3c4\ud569\ub2c8\ub2e4.<\/p>\n
import numpy as np\nimport tensorflow as tf\nfrom tensorflow import keras\nfrom tensorflow.keras import layers\n\n# \uae30\ubcf8\uc801\uc778 \ub370\uc774\ud130 \uc0dd\uc131\ndef create_data(num_samples=1000):\n x = np.random.rand(num_samples, 2)\n y = (x[:, 0] + x[:, 1] > 1).astype(int)\n return x, y\n\n# \ubaa8\ub378 \uc0dd\uc131\ndef create_model():\n model = keras.Sequential([\n layers.Dense(10, activation='relu', input_shape=(2,)),\n layers.Dense(1, activation='sigmoid')\n ])\n model.compile(optimizer='adam', loss='binary_crossentropy', metrics=['accuracy'])\n return model\n\n# \ucd08\uae30 \ub370\uc774\ud130 \ud559\uc2b5\nx1, y1 = create_data()\nmodel = create_model()\nmodel.fit(x1, y1, epochs=10)\n\n# \uc0c8\ub85c\uc6b4 \ub370\uc774\ud130 \uc0dd\uc131 \ubc0f \ud559\uc2b5\nx2, y2 = create_data()\nmodel.layers[0].trainable = False # \uccab \ubc88\uc9f8 \ub808\uc774\uc5b4\uc758 \uac00\uc911\uce58\ub97c \uace0\uc815\nmodel.fit(x2, y2, epochs=10) # \uc0c8\ub85c\uc6b4 \ub370\uc774\ud130 \ud559\uc2b5\n<\/code><\/pre>\n8. \uacb0\ub860<\/h2>\n
\uc9c0\uc18d \ud559\uc2b5\uacfc \ud3c9\uc0dd \ud559\uc2b5\uc740 \ud604\ub300 \uc778\uacf5\uc9c0\ub2a5 \uc2dc\uc2a4\ud15c\uc5d0 \ud544\uc218\uc801\uc778 \uc694\uc18c\ub85c, \ub2e4\uc591\ud55c \ub3c4\uc804 \uacfc\uc81c\ub97c \uc548\uace0 \uc788\uc2b5\ub2c8\ub2e4. \uc774 \uae00\uc5d0\uc11c\ub294 \uc9c0\uc18d \ud559\uc2b5\uc758 \ud544\uc694\uc131\uacfc \ub3c4\uc804 \uacfc\uc81c, \uadf8\ub9ac\uace0 \uac04\ub2e8\ud55c \uc608\uc81c\ub97c \ud1b5\ud574 \uc774 \uac1c\ub150\uc744 \uc18c\uac1c\ud558\uc600\uc2b5\ub2c8\ub2e4. \uc55e\uc73c\ub85c\uc758 \uc5f0\uad6c\ub294 \uc774\ub7ec\ud55c \ubb38\uc81c\ub97c \ud574\uacb0\ud558\uace0 \uc9c0\ub2a5\ud615 \uc2dc\uc2a4\ud15c\uc758 \ubc1c\uc804\uc744 \uc774\ub04c\uc5b4\uac00\ub294 \ub370 \uae30\uc5ec\ud560 \uac83\uc785\ub2c8\ub2e4.<\/p>\n
9. \ucc38\uace0 \uc790\ub8cc<\/h2>\n