From fe5fdb9523c852de7f31183d0b0121759775185f Mon Sep 17 00:00:00 2001 From: visionNoob Date: Wed, 12 Sep 2018 20:04:37 +0900 Subject: [PATCH] #40: Translate 3.6 - paragraph 5, 6 / 9 --- .../3.6-classifying-newswires.ipynb" | 169 ++++++++++++------ 1 file changed, 113 insertions(+), 56 deletions(-) diff --git "a/40_Keras\354\231\200 Jupyter Notebook\354\234\274\353\241\234 \354\213\234\354\236\221\355\225\230\353\212\224 \353\224\245\353\237\254\353\213\235/3.6-classifying-newswires.ipynb" "b/40_Keras\354\231\200 Jupyter Notebook\354\234\274\353\241\234 \354\213\234\354\236\221\355\225\230\353\212\224 \353\224\245\353\237\254\353\213\235/3.6-classifying-newswires.ipynb" index 9425335..fbae844 100644 --- "a/40_Keras\354\231\200 Jupyter Notebook\354\234\274\353\241\234 \354\213\234\354\236\221\355\225\230\353\212\224 \353\224\245\353\237\254\353\213\235/3.6-classifying-newswires.ipynb" +++ "b/40_Keras\354\231\200 Jupyter Notebook\354\234\274\353\241\234 \354\213\234\354\236\221\355\225\230\353\212\224 \353\224\245\353\237\254\353\213\235/3.6-classifying-newswires.ipynb" @@ -572,13 +572,19 @@ "cell_type": "markdown", "metadata": {}, "source": [ - "Now let's train our network for 20 epochs:\n", - "네트워크를 20 에포크 동안 학습시켜봅시다" + "Now let's train our network for 20 epochs:" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "네트워크를 20 에포크 동안 학습시켜봅시다:" ] }, { "cell_type": "code", - "execution_count": 15, + "execution_count": 13, "metadata": {}, "outputs": [ { @@ -587,45 +593,45 @@ "text": [ "Train on 7982 samples, validate on 1000 samples\n", "Epoch 1/20\n", - "7982/7982 [==============================] - 1s - loss: 2.5241 - acc: 0.4952 - val_loss: 1.7263 - val_acc: 0.6100\n", + "7982/7982 [==============================] - 2s 241us/step - loss: 2.5322 - acc: 0.4955 - val_loss: 1.7208 - val_acc: 0.6120\n", "Epoch 2/20\n", - "7982/7982 [==============================] - 0s - loss: 1.4500 - acc: 0.6854 - val_loss: 1.3478 - val_acc: 0.7070\n", + "7982/7982 [==============================] - 1s 101us/step - loss: 1.4452 - acc: 0.6879 - val_loss: 1.3459 - val_acc: 0.7060\n", "Epoch 3/20\n", - "7982/7982 [==============================] - 0s - loss: 1.0979 - acc: 0.7643 - val_loss: 1.1736 - val_acc: 0.7460\n", + "7982/7982 [==============================] - 1s 101us/step - loss: 1.0953 - acc: 0.7651 - val_loss: 1.1708 - val_acc: 0.7430\n", "Epoch 4/20\n", - "7982/7982 [==============================] - 0s - loss: 0.8723 - acc: 0.8178 - val_loss: 1.0880 - val_acc: 0.7490\n", + "7982/7982 [==============================] - 1s 100us/step - loss: 0.8697 - acc: 0.8165 - val_loss: 1.0793 - val_acc: 0.7590\n", "Epoch 5/20\n", - "7982/7982 [==============================] - 0s - loss: 0.7045 - acc: 0.8477 - val_loss: 0.9822 - val_acc: 0.7760\n", + "7982/7982 [==============================] - 1s 101us/step - loss: 0.7034 - acc: 0.8472 - val_loss: 0.9844 - val_acc: 0.7810\n", "Epoch 6/20\n", - "7982/7982 [==============================] - 0s - loss: 0.5660 - acc: 0.8792 - val_loss: 0.9379 - val_acc: 0.8030\n", + "7982/7982 [==============================] - 1s 101us/step - loss: 0.5667 - acc: 0.8802 - val_loss: 0.9411 - val_acc: 0.8040\n", "Epoch 7/20\n", - "7982/7982 [==============================] - 0s - loss: 0.4569 - acc: 0.9037 - val_loss: 0.9039 - val_acc: 0.8050\n", + "7982/7982 [==============================] - 1s 99us/step - loss: 0.4581 - acc: 0.9048 - val_loss: 0.9083 - val_acc: 0.8020\n", "Epoch 8/20\n", - "7982/7982 [==============================] - 0s - loss: 0.3668 - acc: 0.9238 - val_loss: 0.9279 - val_acc: 0.7890\n", + "7982/7982 [==============================] - 1s 102us/step - loss: 0.3695 - acc: 0.9231 - val_loss: 0.9363 - val_acc: 0.7890\n", "Epoch 9/20\n", - "7982/7982 [==============================] - 0s - loss: 0.3000 - acc: 0.9326 - val_loss: 0.8835 - val_acc: 0.8070\n", + "7982/7982 [==============================] - 1s 102us/step - loss: 0.3032 - acc: 0.9315 - val_loss: 0.8917 - val_acc: 0.8090\n", "Epoch 10/20\n", - "7982/7982 [==============================] - 0s - loss: 0.2505 - acc: 0.9434 - val_loss: 0.8967 - val_acc: 0.8150\n", + "7982/7982 [==============================] - 1s 101us/step - loss: 0.2537 - acc: 0.9414 - val_loss: 0.9071 - val_acc: 0.8110\n", "Epoch 11/20\n", - "7982/7982 [==============================] - 0s - loss: 0.2155 - acc: 0.9473 - val_loss: 0.9080 - val_acc: 0.8110\n", + "7982/7982 [==============================] - 1s 99us/step - loss: 0.2187 - acc: 0.9471 - val_loss: 0.9177 - val_acc: 0.8130\n", "Epoch 12/20\n", - "7982/7982 [==============================] - 0s - loss: 0.1853 - acc: 0.9506 - val_loss: 0.9025 - val_acc: 0.8140\n", + "7982/7982 [==============================] - 1s 100us/step - loss: 0.1873 - acc: 0.9508 - val_loss: 0.9027 - val_acc: 0.8130\n", "Epoch 13/20\n", - "7982/7982 [==============================] - 0s - loss: 0.1680 - acc: 0.9524 - val_loss: 0.9268 - val_acc: 0.8100\n", + "7982/7982 [==============================] - 1s 101us/step - loss: 0.1703 - acc: 0.9521 - val_loss: 0.9323 - val_acc: 0.8110\n", "Epoch 14/20\n", - "7982/7982 [==============================] - 0s - loss: 0.1512 - acc: 0.9562 - val_loss: 0.9500 - val_acc: 0.8130\n", + "7982/7982 [==============================] - 1s 101us/step - loss: 0.1536 - acc: 0.9554 - val_loss: 0.9689 - val_acc: 0.8050\n", "Epoch 15/20\n", - "7982/7982 [==============================] - 0s - loss: 0.1371 - acc: 0.9559 - val_loss: 0.9621 - val_acc: 0.8090\n", + "7982/7982 [==============================] - 1s 100us/step - loss: 0.1390 - acc: 0.9560 - val_loss: 0.9686 - val_acc: 0.8150\n", "Epoch 16/20\n", - "7982/7982 [==============================] - 0s - loss: 0.1306 - acc: 0.9553 - val_loss: 1.0152 - val_acc: 0.8050\n", + "7982/7982 [==============================] - 1s 100us/step - loss: 0.1313 - acc: 0.9560 - val_loss: 1.0220 - val_acc: 0.8060\n", "Epoch 17/20\n", - "7982/7982 [==============================] - 0s - loss: 0.1210 - acc: 0.9575 - val_loss: 1.0262 - val_acc: 0.8010\n", + "7982/7982 [==============================] - 1s 101us/step - loss: 0.1217 - acc: 0.9579 - val_loss: 1.0254 - val_acc: 0.7970\n", "Epoch 18/20\n", - "7982/7982 [==============================] - 0s - loss: 0.1185 - acc: 0.9570 - val_loss: 1.0354 - val_acc: 0.8040\n", + "7982/7982 [==============================] - 1s 102us/step - loss: 0.1198 - acc: 0.9582 - val_loss: 1.0430 - val_acc: 0.8060\n", "Epoch 19/20\n", - "7982/7982 [==============================] - 0s - loss: 0.1128 - acc: 0.9598 - val_loss: 1.0841 - val_acc: 0.8010\n", + "7982/7982 [==============================] - 1s 101us/step - loss: 0.1138 - acc: 0.9597 - val_loss: 1.0955 - val_acc: 0.7970\n", "Epoch 20/20\n", - "7982/7982 [==============================] - 0s - loss: 0.1097 - acc: 0.9594 - val_loss: 1.0707 - val_acc: 0.8040\n" + "7982/7982 [==============================] - 1s 101us/step - loss: 0.1111 - acc: 0.9593 - val_loss: 1.0674 - val_acc: 0.8020\n" ] } ], @@ -644,16 +650,23 @@ "Let's display its loss and accuracy curves:" ] }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "손실과 정확도 커브를 출력시켜봅시다." + ] + }, { "cell_type": "code", - "execution_count": 16, + "execution_count": 14, "metadata": {}, "outputs": [ { "data": { - "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYUAAAEWCAYAAACJ0YulAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nEBGRJCUFERFJUlIQEZGk/w+dYzD20tOTDwAAAABJRU5ErkJggg==\n", + "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYUAAAEWCAYAAACJ0YulAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n", "text/plain": [ - "" + "" ] }, "metadata": {}, @@ -680,14 +693,14 @@ }, { "cell_type": "code", - "execution_count": 18, + "execution_count": 15, "metadata": {}, "outputs": [ { "data": { - "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYUAAAEWCAYAAACJ0YulAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nEZEEJQUREUn4/0ypeWoxAVOLAAAAAElFTkSuQmCC\n", + "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYUAAAEWCAYAAACJ0YulAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\npk2ErJsAAAAASUVORK5CYII=\n", "text/plain": [ - "" + "" ] }, "metadata": {}, @@ -718,9 +731,16 @@ "the test set:" ] }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "네트워크가 8에포크 이후로는 과적합하는 것 같습니다. 새로운 네트워크를 처음부터 8에포크 까지만 학습시키고 시험 집합을 평가해봅시다." + ] + }, { "cell_type": "code", - "execution_count": 27, + "execution_count": 16, "metadata": {}, "outputs": [ { @@ -729,22 +749,22 @@ "text": [ "Train on 7982 samples, validate on 1000 samples\n", "Epoch 1/8\n", - "7982/7982 [==============================] - 0s - loss: 2.6118 - acc: 0.4667 - val_loss: 1.7207 - val_acc: 0.6360\n", + "7982/7982 [==============================] - 1s 148us/step - loss: 2.5398 - acc: 0.5226 - val_loss: 1.6733 - val_acc: 0.6570\n", "Epoch 2/8\n", - "7982/7982 [==============================] - 0s - loss: 1.3998 - acc: 0.7107 - val_loss: 1.2645 - val_acc: 0.7360\n", + "7982/7982 [==============================] - 1s 100us/step - loss: 1.3712 - acc: 0.7121 - val_loss: 1.2758 - val_acc: 0.7210\n", "Epoch 3/8\n", - "7982/7982 [==============================] - 0s - loss: 1.0343 - acc: 0.7839 - val_loss: 1.0994 - val_acc: 0.7700\n", + "7982/7982 [==============================] - 1s 101us/step - loss: 1.0136 - acc: 0.7781 - val_loss: 1.1303 - val_acc: 0.7530\n", "Epoch 4/8\n", - "7982/7982 [==============================] - 0s - loss: 0.8114 - acc: 0.8329 - val_loss: 1.0252 - val_acc: 0.7820\n", + "7982/7982 [==============================] - 1s 114us/step - loss: 0.7976 - acc: 0.8251 - val_loss: 1.0539 - val_acc: 0.7590\n", "Epoch 5/8\n", - "7982/7982 [==============================] - 0s - loss: 0.6466 - acc: 0.8628 - val_loss: 0.9536 - val_acc: 0.8070\n", + "7982/7982 [==============================] - 1s 103us/step - loss: 0.6393 - acc: 0.8624 - val_loss: 0.9754 - val_acc: 0.7920\n", "Epoch 6/8\n", - "7982/7982 [==============================] - 0s - loss: 0.5271 - acc: 0.8894 - val_loss: 0.9187 - val_acc: 0.8110\n", + "7982/7982 [==============================] - 1s 100us/step - loss: 0.5124 - acc: 0.8921 - val_loss: 0.9102 - val_acc: 0.8140\n", "Epoch 7/8\n", - "7982/7982 [==============================] - 0s - loss: 0.4193 - acc: 0.9126 - val_loss: 0.9051 - val_acc: 0.8120\n", + "7982/7982 [==============================] - 1s 102us/step - loss: 0.4124 - acc: 0.9137 - val_loss: 0.8932 - val_acc: 0.8210\n", "Epoch 8/8\n", - "7982/7982 [==============================] - 0s - loss: 0.3478 - acc: 0.9258 - val_loss: 0.8891 - val_acc: 0.8160\n", - "1952/2246 [=========================>....] - ETA: 0s" + "7982/7982 [==============================] - 1s 130us/step - loss: 0.3355 - acc: 0.9290 - val_loss: 0.8732 - val_acc: 0.8260\n", + "2246/2246 [==============================] - 0s 192us/step\n" ] } ], @@ -767,16 +787,16 @@ }, { "cell_type": "code", - "execution_count": 28, + "execution_count": 17, "metadata": {}, "outputs": [ { "data": { "text/plain": [ - "[0.98764628548762257, 0.77693677651807869]" + "[0.9845061221509137, 0.7836153161175423]" ] }, - "execution_count": 28, + "execution_count": 17, "metadata": {}, "output_type": "execute_result" } @@ -789,23 +809,30 @@ "cell_type": "markdown", "metadata": {}, "source": [ - "\n", "Our approach reaches an accuracy of ~78%. With a balanced binary classification problem, the accuracy reached by a purely random classifier \n", "would be 50%, but in our case it is closer to 19%, so our results seem pretty good, at least when compared to a random baseline:" ] }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "우리의 모델은 78%의 정확도를 기록했습니다. 데이터 집합이 고루 분포된 이진 분류 문제의 경우, 둘 중 하나를 찍는 분류기의 성능은 적확도는 50% 일 것입니다. \n", + "하지만 우리의 경우는 46 클래스를 분류하는 문제이므로, 아무리 잘 찍어도 19% 를 넘기긴 힘들 것입니다. 따라서 78% 라는 정확도는 아주 좋습니다. 적어도 그냥 찍는 모델에 비해서는 말이죠." + ] + }, { "cell_type": "code", - "execution_count": 29, + "execution_count": 21, "metadata": {}, "outputs": [ { "data": { "text/plain": [ - "0.18477292965271594" + "0.1834372217275156" ] }, - "execution_count": 29, + "execution_count": 21, "metadata": {}, "output_type": "execute_result" } @@ -825,13 +852,22 @@ "# Paragraph 6\n", "## Generating predictions on new data\n", "\n", - "We can verify that the `predict` method of our model instance returns a probability distribution over all 46 topics. Let's generate topic \n", - "predictions for all of the test data:" + "We can verify that the `predict` method of our model instance returns a probability distribution over all 46 topics. Let's generate topic predictions for all of the test data:" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "# Paragraph 6\n", + "## 새로운 데이터에 대한 예측 모델 생성하기 \n", + "\n", + "`predict` 라는 메소드가 46개 토픽에 대한 확률 분포를 반환함을 알 수 있습니다. 자 그럼 모든 시험 데이터에 대해서 토픽을 예측해봅시다. " ] }, { "cell_type": "code", - "execution_count": 30, + "execution_count": 23, "metadata": { "collapsed": true }, @@ -847,9 +883,16 @@ "Each entry in `predictions` is a vector of length 46:" ] }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "`prediction` 변수의 각 요소는 길이가 46인 벡터입니다:" + ] + }, { "cell_type": "code", - "execution_count": 31, + "execution_count": 24, "metadata": {}, "outputs": [ { @@ -858,7 +901,7 @@ "(46,)" ] }, - "execution_count": 31, + "execution_count": 24, "metadata": {}, "output_type": "execute_result" } @@ -874,18 +917,25 @@ "The coefficients in this vector sum to 1:" ] }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "이 벡터의 각 요소를 더하면 1이 됩니다:" + ] + }, { "cell_type": "code", - "execution_count": 32, + "execution_count": 25, "metadata": {}, "outputs": [ { "data": { "text/plain": [ - "0.99999994" + "1.0000001" ] }, - "execution_count": 32, + "execution_count": 25, "metadata": {}, "output_type": "execute_result" } @@ -901,9 +951,16 @@ "The largest entry is the predicted class, i.e. the class with the highest probability:" ] }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "그리고 그 중 가장 값이 높은 클래스로 예측합니다. 즉 가장 확률이 높은 클래스를 선택하는 것입니다." + ] + }, { "cell_type": "code", - "execution_count": 33, + "execution_count": 26, "metadata": {}, "outputs": [ { @@ -912,7 +969,7 @@ "3" ] }, - "execution_count": 33, + "execution_count": 26, "metadata": {}, "output_type": "execute_result" }