chapter03_introduction-to-keras-and-tf.ipynb

{
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "LfldcZzRjR1L"
      },
      "source": [
        "이 노트북은 [케라스 창시자에게 배우는 딥러닝 2판](https://tensorflow.blog/kerasdl2/)의 예제 코드를 담고 있습니다.\n",
        "\n",
        "<table align=\"left\"><tr><td>\n",
        "<a href=\"https://colab.research.google.com/github/rickiepark/deep-learning-with-python-2nd/blob/master/chapter03_introduction-to-keras-and-tf.ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a>\n",
        "</td></tr></table>"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "iKORndCjjR1N"
      },
      "source": [
        "# 케라스와 텐서플로 소개"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "Bnvrkq6pjR1O"
      },
      "source": [
        "## 텐서플로란?"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "gXa_qKgAjR1O"
      },
      "source": [
        "## 케라스란?"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "9Ti2lq6PjR1O"
      },
      "source": [
        "## 케라스와 텐서플로의 간략한 역사"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "RS5aSY-JjR1O"
      },
      "source": [
        "## 딥러닝 작업 환경 설정하기"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "tpAPYi2NjR1P"
      },
      "source": [
        "### 주피터 노트북: 권장하는 딥러닝 실험 도구"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "j0JRUfctjR1P"
      },
      "source": [
        "### 코랩 사용하기"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "S99ikTFVjR1Q"
      },
      "source": [
        "#### 코랩 시작하기"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "X58JBJtljR1Q"
      },
      "source": [
        "#### pip로 패키지 설치하기"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "dCnG27L6jR1R"
      },
      "source": [
        "#### GPU 런타임 사용하기"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "djDuvOKVjR1R"
      },
      "source": [
        "## 텐서플로 시작하기"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "T887Zne6jR1R"
      },
      "source": [
        "#### 상수 텐서와 변수"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "e8ykr8bMjR1R"
      },
      "source": [
        "**모두 1 또는 모두 0인 텐서**"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 1,
      "metadata": {
        "execution": {
          "iopub.execute_input": "2022-05-21T00:58:14.681100Z",
          "iopub.status.busy": "2022-05-21T00:58:14.680540Z",
          "iopub.status.idle": "2022-05-21T00:58:16.481810Z",
          "shell.execute_reply": "2022-05-21T00:58:16.480769Z"
        },
        "id": "0Yh_78IzjR1R",
        "outputId": "ecf11ea3-2112-4e27-d93c-38d10cb521b0",
        "colab": {
          "base_uri": "https://localhost:8080/"
        }
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "tf.Tensor(\n",
            "[[1.]\n",
            " [1.]], shape=(2, 1), dtype=float32)\n"
          ]
        }
      ],
      "source": [
        "import tensorflow as tf\n",
        "x = tf.ones(shape=(2, 1))\n",
        "print(x)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 2,
      "metadata": {
        "execution": {
          "iopub.execute_input": "2022-05-21T00:58:16.485729Z",
          "iopub.status.busy": "2022-05-21T00:58:16.485088Z",
          "iopub.status.idle": "2022-05-21T00:58:16.490852Z",
          "shell.execute_reply": "2022-05-21T00:58:16.490149Z"
        },
        "id": "63xkTvoXjR1T",
        "outputId": "46fa3899-9a70-41b7-9361-f169a8633452",
        "colab": {
          "base_uri": "https://localhost:8080/"
        }
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "tf.Tensor(\n",
            "[[0.]\n",
            " [0.]], shape=(2, 1), dtype=float32)\n"
          ]
        }
      ],
      "source": [
        "x = tf.zeros(shape=(2, 1))\n",
        "print(x)"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "NthSViHdjR1T"
      },
      "source": [
        "**랜덤 텐서**"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 3,
      "metadata": {
        "execution": {
          "iopub.execute_input": "2022-05-21T00:58:16.494374Z",
          "iopub.status.busy": "2022-05-21T00:58:16.493742Z",
          "iopub.status.idle": "2022-05-21T00:58:16.501220Z",
          "shell.execute_reply": "2022-05-21T00:58:16.500464Z"
        },
        "id": "Ktevr051jR1T",
        "outputId": "8b50b09d-ab92-4fb7-c3d8-7809dca4c275",
        "colab": {
          "base_uri": "https://localhost:8080/"
        }
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "tf.Tensor(\n",
            "[[ 0.6189793 ]\n",
            " [ 0.77359086]\n",
            " [-0.66645133]], shape=(3, 1), dtype=float32)\n"
          ]
        }
      ],
      "source": [
        "x = tf.random.normal(shape=(3, 1), mean=0., stddev=1.)\n",
        "print(x)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 4,
      "metadata": {
        "execution": {
          "iopub.execute_input": "2022-05-21T00:58:16.504901Z",
          "iopub.status.busy": "2022-05-21T00:58:16.504410Z",
          "iopub.status.idle": "2022-05-21T00:58:16.510480Z",
          "shell.execute_reply": "2022-05-21T00:58:16.509458Z"
        },
        "id": "lcm6ziA-jR1T",
        "outputId": "560c99b8-e33c-4ff4-91b8-4d370a0fab06",
        "colab": {
          "base_uri": "https://localhost:8080/"
        }
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "tf.Tensor(\n",
            "[[0.05876148]\n",
            " [0.3948239 ]\n",
            " [0.3419211 ]], shape=(3, 1), dtype=float32)\n"
          ]
        }
      ],
      "source": [
        "x = tf.random.uniform(shape=(3, 1), minval=0., maxval=1.)\n",
        "print(x)"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "Xr-7w0jHjR1U"
      },
      "source": [
        "**넘파이 배열에 값 할당하기**"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 5,
      "metadata": {
        "execution": {
          "iopub.execute_input": "2022-05-21T00:58:16.514734Z",
          "iopub.status.busy": "2022-05-21T00:58:16.514199Z",
          "iopub.status.idle": "2022-05-21T00:58:16.518209Z",
          "shell.execute_reply": "2022-05-21T00:58:16.517567Z"
        },
        "id": "0HnGGxzJjR1U"
      },
      "outputs": [],
      "source": [
        "import numpy as np\n",
        "x = np.ones(shape=(2, 2))\n",
        "x[0, 0] = 0."
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "exaFII_wjR1U"
      },
      "source": [
        "**텐서플로 변수 만들기**"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 6,
      "metadata": {
        "execution": {
          "iopub.execute_input": "2022-05-21T00:58:16.522164Z",
          "iopub.status.busy": "2022-05-21T00:58:16.521381Z",
          "iopub.status.idle": "2022-05-21T00:58:16.529208Z",
          "shell.execute_reply": "2022-05-21T00:58:16.528387Z"
        },
        "id": "W3dtiJ7hjR1U",
        "outputId": "bb0daa21-e1e6-4d49-cbb4-a8944e98d88c",
        "colab": {
          "base_uri": "https://localhost:8080/"
        }
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "<tf.Variable 'Variable:0' shape=(3, 1) dtype=float32, numpy=\n",
            "array([[-0.27607322],\n",
            "       [-2.4914272 ],\n",
            "       [ 1.2171313 ]], dtype=float32)>\n"
          ]
        }
      ],
      "source": [
        "v = tf.Variable(initial_value=tf.random.normal(shape=(3, 1)))\n",
        "print(v)"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "ylun-78QjR1V"
      },
      "source": [
        "**텐서플로 변수에 값 할당하기**"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 7,
      "metadata": {
        "execution": {
          "iopub.execute_input": "2022-05-21T00:58:16.532960Z",
          "iopub.status.busy": "2022-05-21T00:58:16.532657Z",
          "iopub.status.idle": "2022-05-21T00:58:16.545443Z",
          "shell.execute_reply": "2022-05-21T00:58:16.544477Z"
        },
        "id": "3L7pyEarjR1V",
        "outputId": "7faa42ff-2e16-4f31-f562-707f6891a286",
        "colab": {
          "base_uri": "https://localhost:8080/"
        }
      },
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<tf.Variable 'UnreadVariable' shape=(3, 1) dtype=float32, numpy=\n",
              "array([[1.],\n",
              "       [1.],\n",
              "       [1.]], dtype=float32)>"
            ]
          },
          "metadata": {},
          "execution_count": 7
        }
      ],
      "source": [
        "v.assign(tf.ones((3, 1)))"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "slQzosWHjR1V"
      },
      "source": [
        "**변수 일부에 값 할당하기**"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 8,
      "metadata": {
        "execution": {
          "iopub.execute_input": "2022-05-21T00:58:16.549025Z",
          "iopub.status.busy": "2022-05-21T00:58:16.548733Z",
          "iopub.status.idle": "2022-05-21T00:58:16.557519Z",
          "shell.execute_reply": "2022-05-21T00:58:16.556482Z"
        },
        "id": "RM9JbVFMjR1V",
        "outputId": "e9039823-f55c-4473-cbfd-4a2bcd2d7185",
        "colab": {
          "base_uri": "https://localhost:8080/"
        }
      },
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<tf.Variable 'UnreadVariable' shape=(3, 1) dtype=float32, numpy=\n",
              "array([[3.],\n",
              "       [1.],\n",
              "       [1.]], dtype=float32)>"
            ]
          },
          "metadata": {},
          "execution_count": 8
        }
      ],
      "source": [
        "v[0, 0].assign(3.)"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "m2PtAFsgjR1V"
      },
      "source": [
        "**`assign_add` 사용하기**"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 9,
      "metadata": {
        "execution": {
          "iopub.execute_input": "2022-05-21T00:58:16.561236Z",
          "iopub.status.busy": "2022-05-21T00:58:16.560971Z",
          "iopub.status.idle": "2022-05-21T00:58:16.568357Z",
          "shell.execute_reply": "2022-05-21T00:58:16.567523Z"
        },
        "id": "nRvjF0NKjR1W",
        "outputId": "9257b875-0ded-4403-ff38-3a1e78a7d0dd",
        "colab": {
          "base_uri": "https://localhost:8080/"
        }
      },
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<tf.Variable 'UnreadVariable' shape=(3, 1) dtype=float32, numpy=\n",
              "array([[4.],\n",
              "       [2.],\n",
              "       [2.]], dtype=float32)>"
            ]
          },
          "metadata": {},
          "execution_count": 9
        }
      ],
      "source": [
        "v.assign_add(tf.ones((3, 1)))"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "ZdevQ8z-jR1W"
      },
      "source": [
        "#### 텐서 연산: 텐서플로에서 수학 계산하기"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "tckVPX4tjR1W"
      },
      "source": [
        "**기본적인 수학 연산**"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 10,
      "metadata": {
        "execution": {
          "iopub.execute_input": "2022-05-21T00:58:16.572052Z",
          "iopub.status.busy": "2022-05-21T00:58:16.571786Z",
          "iopub.status.idle": "2022-05-21T00:58:16.599467Z",
          "shell.execute_reply": "2022-05-21T00:58:16.598533Z"
        },
        "id": "2cfJ4xb9jR1W"
      },
      "outputs": [],
      "source": [
        "a = tf.ones((2, 2))\n",
        "b = tf.square(a)\n",
        "c = tf.sqrt(a)\n",
        "d = b + c\n",
        "e = tf.matmul(a, b)\n",
        "e *= d"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "yD4QBR3LjR1W"
      },
      "source": [
        "#### GradientTape API 다시 살펴 보기"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "wf9L61oejR1W"
      },
      "source": [
        "**`GradientTape` 사용하기**"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 11,
      "metadata": {
        "execution": {
          "iopub.execute_input": "2022-05-21T00:58:16.603842Z",
          "iopub.status.busy": "2022-05-21T00:58:16.603513Z",
          "iopub.status.idle": "2022-05-21T00:58:16.612102Z",
          "shell.execute_reply": "2022-05-21T00:58:16.611176Z"
        },
        "id": "r2NyvJv1jR1W"
      },
      "outputs": [],
      "source": [
        "input_var = tf.Variable(initial_value=3.)\n",
        "with tf.GradientTape() as tape:\n",
        "   result = tf.square(input_var)\n",
        "gradient = tape.gradient(result, input_var)"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "o9CCUOnqjR1W"
      },
      "source": [
        "**상수 입력 텐서와 함께 `GradientTape` 사용하기**"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 12,
      "metadata": {
        "execution": {
          "iopub.execute_input": "2022-05-21T00:58:16.615692Z",
          "iopub.status.busy": "2022-05-21T00:58:16.615413Z",
          "iopub.status.idle": "2022-05-21T00:58:16.620532Z",
          "shell.execute_reply": "2022-05-21T00:58:16.619752Z"
        },
        "id": "GeWtkJ5ijR1X"
      },
      "outputs": [],
      "source": [
        "input_const = tf.constant(3.)\n",
        "with tf.GradientTape() as tape:\n",
        "   tape.watch(input_const)\n",
        "   result = tf.square(input_const)\n",
        "gradient = tape.gradient(result, input_const)"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "UOoe3Zp8jR1X"
      },
      "source": [
        "**그레이디언트 테이프를 중첩하여 이계도 그레이디언트를 계산하기**"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 13,
      "metadata": {
        "execution": {
          "iopub.execute_input": "2022-05-21T00:58:16.624217Z",
          "iopub.status.busy": "2022-05-21T00:58:16.623600Z",
          "iopub.status.idle": "2022-05-21T00:58:16.630991Z",
          "shell.execute_reply": "2022-05-21T00:58:16.629989Z"
        },
        "id": "e5gXa8SVjR1X"
      },
      "outputs": [],
      "source": [
        "time = tf.Variable(0.)\n",
        "with tf.GradientTape() as outer_tape:\n",
        "    with tf.GradientTape() as inner_tape:\n",
        "        position =  4.9 * time ** 2\n",
        "    speed = inner_tape.gradient(position, time)\n",
        "acceleration = outer_tape.gradient(speed, time)"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "A1MVfacbjR1X"
      },
      "source": [
        "#### 엔드-투-엔드 예제: 텐서플로 선형 분류기"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "l8RNP89rjR1X"
      },
      "source": [
        "**2D 평면에 두 클래스의 랜덤한 포인트 생성하기**"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 14,
      "metadata": {
        "execution": {
          "iopub.execute_input": "2022-05-21T00:58:16.635105Z",
          "iopub.status.busy": "2022-05-21T00:58:16.634828Z",
          "iopub.status.idle": "2022-05-21T00:58:16.641379Z",
          "shell.execute_reply": "2022-05-21T00:58:16.640512Z"
        },
        "id": "ukRDkZjcjR1X"
      },
      "outputs": [],
      "source": [
        "num_samples_per_class = 1000\n",
        "negative_samples = np.random.multivariate_normal(\n",
        "    mean=[0, 3],\n",
        "    cov=[[1, 0.5],[0.5, 1]],\n",
        "    size=num_samples_per_class)\n",
        "positive_samples = np.random.multivariate_normal(\n",
        "    mean=[3, 0],\n",
        "    cov=[[1, 0.5],[0.5, 1]],\n",
        "    size=num_samples_per_class)"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "XNW_TQz7jR1X"
      },
      "source": [
        "**두 클래스를 (2000, 2) 크기의 한 배열로 쌓기**"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 15,
      "metadata": {
        "execution": {
          "iopub.execute_input": "2022-05-21T00:58:16.645184Z",
          "iopub.status.busy": "2022-05-21T00:58:16.644657Z",
          "iopub.status.idle": "2022-05-21T00:58:16.649566Z",
          "shell.execute_reply": "2022-05-21T00:58:16.648645Z"
        },
        "id": "mKsVyGg0jR1Y"
      },
      "outputs": [],
      "source": [
        "inputs = np.vstack((negative_samples, positive_samples)).astype(np.float32)"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "JR70BhDujR1Y"
      },
      "source": [
        "**(0과 1로 구성된) 타깃 생성하기**"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 16,
      "metadata": {
        "execution": {
          "iopub.execute_input": "2022-05-21T00:58:16.653376Z",
          "iopub.status.busy": "2022-05-21T00:58:16.652893Z",
          "iopub.status.idle": "2022-05-21T00:58:16.656858Z",
          "shell.execute_reply": "2022-05-21T00:58:16.656239Z"
        },
        "id": "jMEgZXkMjR1Y"
      },
      "outputs": [],
      "source": [
        "targets = np.vstack((np.zeros((num_samples_per_class, 1), dtype=\"float32\"),\n",
        "                     np.ones((num_samples_per_class, 1), dtype=\"float32\")))"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "3joL3w9qjR1Y"
      },
      "source": [
        "**두 클래스의 포인트를 그래프로 그리기**"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 17,
      "metadata": {
        "execution": {
          "iopub.execute_input": "2022-05-21T00:58:16.660370Z",
          "iopub.status.busy": "2022-05-21T00:58:16.659928Z",
          "iopub.status.idle": "2022-05-21T00:58:17.496563Z",
          "shell.execute_reply": "2022-05-21T00:58:17.495565Z"
        },
        "id": "9TXfBsUJjR1Y",
        "outputId": "2442cc8e-1ee0-47fd-bf4c-ae77653fd625",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 265
        }
      },
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ],
      "source": [
        "import matplotlib.pyplot as plt\n",
        "plt.scatter(inputs[:, 0], inputs[:, 1], c=targets[:, 0])\n",
        "plt.show()"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "2fZlBGjfjR1Y"
      },
      "source": [
        "**선형 분류기의 변수 만들기**"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 18,
      "metadata": {
        "execution": {
          "iopub.execute_input": "2022-05-21T00:58:17.501331Z",
          "iopub.status.busy": "2022-05-21T00:58:17.500625Z",
          "iopub.status.idle": "2022-05-21T00:58:17.507479Z",
          "shell.execute_reply": "2022-05-21T00:58:17.506630Z"
        },
        "id": "Lr6-tTFtjR1Y"
      },
      "outputs": [],
      "source": [
        "input_dim = 2\n",
        "output_dim = 1\n",
        "W = tf.Variable(initial_value=tf.random.uniform(shape=(input_dim, output_dim)))\n",
        "b = tf.Variable(initial_value=tf.zeros(shape=(output_dim,)))"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "nepUS0hcjR1Y"
      },
      "source": [
        "**정방향 패스 함수**"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 19,
      "metadata": {
        "execution": {
          "iopub.execute_input": "2022-05-21T00:58:17.511530Z",
          "iopub.status.busy": "2022-05-21T00:58:17.510880Z",
          "iopub.status.idle": "2022-05-21T00:58:17.514807Z",
          "shell.execute_reply": "2022-05-21T00:58:17.513996Z"
        },
        "id": "NENOsp6wjR1Z"
      },
      "outputs": [],
      "source": [
        "def model(inputs):\n",
        "    return tf.matmul(inputs, W) + b"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "PS4bb2IFjR1Z"
      },
      "source": [
        "**평균 제곱 오차 손실 함수**"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 20,
      "metadata": {
        "execution": {
          "iopub.execute_input": "2022-05-21T00:58:17.518429Z",
          "iopub.status.busy": "2022-05-21T00:58:17.518140Z",
          "iopub.status.idle": "2022-05-21T00:58:17.522316Z",
          "shell.execute_reply": "2022-05-21T00:58:17.521551Z"
        },
        "id": "TfRT4wnCjR1Z"
      },
      "outputs": [],
      "source": [
        "def square_loss(targets, predictions):\n",
        "    per_sample_losses = tf.square(targets - predictions)\n",
        "    return tf.reduce_mean(per_sample_losses)"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "BnMMSTVojR1a"
      },
      "source": [
        "**훈련 스텝 함수**"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 21,
      "metadata": {
        "execution": {
          "iopub.execute_input": "2022-05-21T00:58:17.526391Z",
          "iopub.status.busy": "2022-05-21T00:58:17.525809Z",
          "iopub.status.idle": "2022-05-21T00:58:17.531136Z",
          "shell.execute_reply": "2022-05-21T00:58:17.530318Z"
        },
        "id": "pfnfnGcbjR1a"
      },
      "outputs": [],
      "source": [
        "learning_rate = 0.1\n",
        "\n",
        "def training_step(inputs, targets):\n",
        "    with tf.GradientTape() as tape:\n",
        "        predictions = model(inputs)\n",
        "        loss = square_loss(targets, predictions)\n",
        "    grad_loss_wrt_W, grad_loss_wrt_b = tape.gradient(loss, [W, b])\n",
        "    W.assign_sub(grad_loss_wrt_W * learning_rate)\n",
        "    b.assign_sub(grad_loss_wrt_b * learning_rate)\n",
        "    return loss"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "AdzwgKLIjR1a"
      },
      "source": [
        "**배치 훈련 루프**"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 22,
      "metadata": {
        "execution": {
          "iopub.execute_input": "2022-05-21T00:58:17.535011Z",
          "iopub.status.busy": "2022-05-21T00:58:17.534413Z",
          "iopub.status.idle": "2022-05-21T00:58:17.601552Z",
          "shell.execute_reply": "2022-05-21T00:58:17.600577Z"
        },
        "id": "EXeU2olUjR1a",
        "outputId": "28f04b28-9bc0-40ca-f0f7-4858938da762",
        "colab": {
          "base_uri": "https://localhost:8080/"
        }
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "0번째 스텝의 손실: 6.7348\n",
            "1번째 스텝의 손실: 1.0284\n",
            "2번째 스텝의 손실: 0.2843\n",
            "3번째 스텝의 손실: 0.1757\n",
            "4번째 스텝의 손실: 0.1508\n",
            "5번째 스텝의 손실: 0.1378\n",
            "6번째 스텝의 손실: 0.1271\n",
            "7번째 스텝의 손실: 0.1177\n",
            "8번째 스텝의 손실: 0.1091\n",
            "9번째 스텝의 손실: 0.1013\n",
            "10번째 스텝의 손실: 0.0942\n",
            "11번째 스텝의 손실: 0.0877\n",
            "12번째 스텝의 손실: 0.0819\n",
            "13번째 스텝의 손실: 0.0766\n",
            "14번째 스텝의 손실: 0.0717\n",
            "15번째 스텝의 손실: 0.0673\n",
            "16번째 스텝의 손실: 0.0633\n",
            "17번째 스텝의 손실: 0.0597\n",
            "18번째 스텝의 손실: 0.0564\n",
            "19번째 스텝의 손실: 0.0534\n",
            "20번째 스텝의 손실: 0.0507\n",
            "21번째 스텝의 손실: 0.0482\n",
            "22번째 스텝의 손실: 0.0460\n",
            "23번째 스텝의 손실: 0.0439\n",
            "24번째 스텝의 손실: 0.0421\n",
            "25번째 스텝의 손실: 0.0404\n",
            "26번째 스텝의 손실: 0.0389\n",
            "27번째 스텝의 손실: 0.0375\n",
            "28번째 스텝의 손실: 0.0362\n",
            "29번째 스텝의 손실: 0.0351\n",
            "30번째 스텝의 손실: 0.0340\n",
            "31번째 스텝의 손실: 0.0331\n",
            "32번째 스텝의 손실: 0.0322\n",
            "33번째 스텝의 손실: 0.0314\n",
            "34번째 스텝의 손실: 0.0307\n",
            "35번째 스텝의 손실: 0.0301\n",
            "36번째 스텝의 손실: 0.0295\n",
            "37번째 스텝의 손실: 0.0289\n",
            "38번째 스텝의 손실: 0.0285\n",
            "39번째 스텝의 손실: 0.0280\n"
          ]
        }
      ],
      "source": [
        "for step in range(40):\n",
        "    loss = training_step(inputs, targets)\n",
        "    print(f\"{step}번째 스텝의 손실: {loss:.4f}\")"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 23,
      "metadata": {
        "execution": {
          "iopub.execute_input": "2022-05-21T00:58:17.605276Z",
          "iopub.status.busy": "2022-05-21T00:58:17.604747Z",
          "iopub.status.idle": "2022-05-21T00:58:17.748955Z",
          "shell.execute_reply": "2022-05-21T00:58:17.746773Z"
        },
        "id": "JR1aqLU5jR1b",
        "outputId": "1717c6d8-cf64-494f-d8e7-47814246a234",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 265
        }
      },
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ],
      "source": [
        "predictions = model(inputs)\n",
        "plt.scatter(inputs[:, 0], inputs[:, 1], c=predictions[:, 0] > 0.5)\n",
        "plt.show()"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 24,
      "metadata": {
        "execution": {
          "iopub.execute_input": "2022-05-21T00:58:17.753051Z",
          "iopub.status.busy": "2022-05-21T00:58:17.752361Z",
          "iopub.status.idle": "2022-05-21T00:58:17.897123Z",
          "shell.execute_reply": "2022-05-21T00:58:17.895862Z"
        },
        "id": "iJrHw9WSjR1b",
        "outputId": "c30ab7f6-995d-48f1-a595-ae10e3a18e7c",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 265
        }
      },
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ],
      "source": [
        "x = np.linspace(-1, 4, 100)\n",
        "# 사실 100개의 x 축 좌표를 만들 필요 없이 시작과 종료 위치만 있어도 됩니다.\n",
        "# x = [-1, 4]\n",
        "y = - W[0] /  W[1] * x + (0.5 - b) / W[1]\n",
        "plt.plot(x, y, \"-r\")\n",
        "plt.scatter(inputs[:, 0], inputs[:, 1], c=predictions[:, 0] > 0.5)\n",
        "plt.show()"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "6Z2drwSqjR1b"
      },
      "source": [
        "## 신경망의 구조: 핵심 Keras API 이해하기"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "tOk_UcQyjR1b"
      },
      "source": [
        "### 층: 딥러닝의 구성 요소"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "R2XTNDbxjR1b"
      },
      "source": [
        "#### 케라스의 Layer 클래스"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "VCGiAeb3jR1b"
      },
      "source": [
        "**`Layer`의 서브클래스로 구현한 `Dense` 층**"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 25,
      "metadata": {
        "execution": {
          "iopub.execute_input": "2022-05-21T00:58:17.901835Z",
          "iopub.status.busy": "2022-05-21T00:58:17.901476Z",
          "iopub.status.idle": "2022-05-21T00:58:17.908604Z",
          "shell.execute_reply": "2022-05-21T00:58:17.907592Z"
        },
        "id": "0gwFDo70jR1b"
      },
      "outputs": [],
      "source": [
        "from tensorflow import keras\n",
        "\n",
        "class SimpleDense(keras.layers.Layer):\n",
        "\n",
        "    def __init__(self, units, activation=None):\n",
        "        super().__init__()\n",
        "        self.units = units\n",
        "        self.activation = activation\n",
        "\n",
        "    def build(self, input_shape):\n",
        "        input_dim = input_shape[-1]\n",
        "        self.W = self.add_weight(shape=(input_dim, self.units),\n",
        "                                 initializer=\"random_normal\")\n",
        "        self.b = self.add_weight(shape=(self.units,),\n",
        "                                 initializer=\"zeros\")\n",
        "\n",
        "    def call(self, inputs):\n",
        "        y = tf.matmul(inputs, self.W) + self.b\n",
        "        if self.activation is not None:\n",
        "            y = self.activation(y)\n",
        "        return y"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 26,
      "metadata": {
        "execution": {
          "iopub.execute_input": "2022-05-21T00:58:17.912833Z",
          "iopub.status.busy": "2022-05-21T00:58:17.912083Z",
          "iopub.status.idle": "2022-05-21T00:58:17.923080Z",
          "shell.execute_reply": "2022-05-21T00:58:17.922276Z"
        },
        "id": "Z2KTF3FijR1c",
        "outputId": "1abcc42b-b145-4646-e148-cb9be937fad5",
        "colab": {
          "base_uri": "https://localhost:8080/"
        }
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "(2, 32)\n"
          ]
        }
      ],
      "source": [
        "my_dense = SimpleDense(units=32, activation=tf.nn.relu)\n",
        "input_tensor = tf.ones(shape=(2, 784))\n",
        "output_tensor = my_dense(input_tensor)\n",
        "print(output_tensor.shape)"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "-VnPv_4djR1c"
      },
      "source": [
        "#### 자동 크기 추론: 동적으로 층 만들기"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 27,
      "metadata": {
        "execution": {
          "iopub.execute_input": "2022-05-21T00:58:17.926464Z",
          "iopub.status.busy": "2022-05-21T00:58:17.926171Z",
          "iopub.status.idle": "2022-05-21T00:58:17.933973Z",
          "shell.execute_reply": "2022-05-21T00:58:17.932941Z"
        },
        "id": "uO_ctIKmjR1c"
      },
      "outputs": [],
      "source": [
        "from tensorflow.keras import layers\n",
        "layer = layers.Dense(32, activation=\"relu\")"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 28,
      "metadata": {
        "execution": {
          "iopub.execute_input": "2022-05-21T00:58:17.937598Z",
          "iopub.status.busy": "2022-05-21T00:58:17.937320Z",
          "iopub.status.idle": "2022-05-21T00:58:17.945965Z",
          "shell.execute_reply": "2022-05-21T00:58:17.944897Z"
        },
        "id": "MZxM9YAxjR1c"
      },
      "outputs": [],
      "source": [
        "from tensorflow.keras import models\n",
        "from tensorflow.keras import layers\n",
        "model = models.Sequential([\n",
        "    layers.Dense(32, activation=\"relu\"),\n",
        "    layers.Dense(32)\n",
        "])"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 29,
      "metadata": {
        "execution": {
          "iopub.execute_input": "2022-05-21T00:58:17.949491Z",
          "iopub.status.busy": "2022-05-21T00:58:17.949222Z",
          "iopub.status.idle": "2022-05-21T00:58:17.957588Z",
          "shell.execute_reply": "2022-05-21T00:58:17.956818Z"
        },
        "id": "7x4nViHrjR1c"
      },
      "outputs": [],
      "source": [
        "model = keras.Sequential([\n",
        "    SimpleDense(32, activation=\"relu\"),\n",
        "    SimpleDense(64, activation=\"relu\"),\n",
        "    SimpleDense(32, activation=\"relu\"),\n",
        "    SimpleDense(10, activation=\"softmax\")\n",
        "])"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "AnOIxT_pjR1c"
      },
      "source": [
        "### 층에서 모델로"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "GuWb-UATjR1c"
      },
      "source": [
        "### “컴파일” 단계: 학습 과정 설정"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 30,
      "metadata": {
        "execution": {
          "iopub.execute_input": "2022-05-21T00:58:17.961340Z",
          "iopub.status.busy": "2022-05-21T00:58:17.961064Z",
          "iopub.status.idle": "2022-05-21T00:58:17.973759Z",
          "shell.execute_reply": "2022-05-21T00:58:17.972882Z"
        },
        "id": "a6R5ru-3jR1c"
      },
      "outputs": [],
      "source": [
        "model = keras.Sequential([keras.layers.Dense(1)])\n",
        "model.compile(optimizer=\"rmsprop\",\n",
        "              loss=\"mean_squared_error\",\n",
        "              metrics=[\"accuracy\"])"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 31,
      "metadata": {
        "execution": {
          "iopub.execute_input": "2022-05-21T00:58:17.977477Z",
          "iopub.status.busy": "2022-05-21T00:58:17.977176Z",
          "iopub.status.idle": "2022-05-21T00:58:17.989690Z",
          "shell.execute_reply": "2022-05-21T00:58:17.988911Z"
        },
        "id": "uA44NTPejR1c"
      },
      "outputs": [],
      "source": [
        "model.compile(optimizer=keras.optimizers.RMSprop(),\n",
        "              loss=keras.losses.MeanSquaredError(),\n",
        "              metrics=[keras.metrics.BinaryAccuracy()])"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "3lnKn8lZjR1d"
      },
      "source": [
        "### 손실 함수 선택하기"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "cDOSr9pXjR1d"
      },
      "source": [
        "### fit() 메서드 이해하기"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "tmBftilFjR1d"
      },
      "source": [
        "**넘파이 데이터로 `fit()` 메서드 호출하기**"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 32,
      "metadata": {
        "execution": {
          "iopub.execute_input": "2022-05-21T00:58:17.993795Z",
          "iopub.status.busy": "2022-05-21T00:58:17.993504Z",
          "iopub.status.idle": "2022-05-21T00:58:18.510003Z",
          "shell.execute_reply": "2022-05-21T00:58:18.509248Z"
        },
        "id": "TCD2GNU2jR1d",
        "outputId": "92647832-705d-4bbc-e648-f5f92ced9f6c",
        "colab": {
          "base_uri": "https://localhost:8080/"
        }
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Epoch 1/5\n",
            "16/16 [==============================] - 1s 2ms/step - loss: 10.1885 - binary_accuracy: 0.0020\n",
            "Epoch 2/5\n",
            "16/16 [==============================] - 0s 2ms/step - loss: 9.8145 - binary_accuracy: 0.0020 \n",
            "Epoch 3/5\n",
            "16/16 [==============================] - 0s 2ms/step - loss: 9.4990 - binary_accuracy: 0.0020\n",
            "Epoch 4/5\n",
            "16/16 [==============================] - 0s 2ms/step - loss: 9.1918 - binary_accuracy: 0.0020\n",
            "Epoch 5/5\n",
            "16/16 [==============================] - 0s 1ms/step - loss: 8.8920 - binary_accuracy: 0.0020\n"
          ]
        }
      ],
      "source": [
        "history = model.fit(\n",
        "    inputs,\n",
        "    targets,\n",
        "    epochs=5,\n",
        "    batch_size=128\n",
        ")"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 33,
      "metadata": {
        "execution": {
          "iopub.execute_input": "2022-05-21T00:58:18.513804Z",
          "iopub.status.busy": "2022-05-21T00:58:18.513289Z",
          "iopub.status.idle": "2022-05-21T00:58:18.518984Z",
          "shell.execute_reply": "2022-05-21T00:58:18.518076Z"
        },
        "id": "7g0R_STajR1d",
        "outputId": "3c062060-d34a-4782-e19e-08338b3a4224",
        "colab": {
          "base_uri": "https://localhost:8080/"
        }
      },
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "{'loss': [10.188525199890137,\n",
              "  9.814451217651367,\n",
              "  9.498977661132812,\n",
              "  9.191787719726562,\n",
              "  8.891997337341309],\n",
              " 'binary_accuracy': [0.0020000000949949026,\n",
              "  0.0020000000949949026,\n",
              "  0.0020000000949949026,\n",
              "  0.0020000000949949026,\n",
              "  0.0020000000949949026]}"
            ]
          },
          "metadata": {},
          "execution_count": 33
        }
      ],
      "source": [
        "history.history"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "eg7A4QwCjR1d"
      },
      "source": [
        "### 검증 데이터에서 손실과 측정 지표 모니터링하기"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "IcScui7rjR1d"
      },
      "source": [
        "**`validation_data` 매개변수 사용하기**"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 34,
      "metadata": {
        "execution": {
          "iopub.execute_input": "2022-05-21T00:58:18.522664Z",
          "iopub.status.busy": "2022-05-21T00:58:18.522375Z",
          "iopub.status.idle": "2022-05-21T00:58:19.764061Z",
          "shell.execute_reply": "2022-05-21T00:58:19.763028Z"
        },
        "id": "KaP6UYfNjR1d",
        "outputId": "02f2ad49-f255-43c3-cf33-30d32089b7bd",
        "colab": {
          "base_uri": "https://localhost:8080/"
        }
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Epoch 1/5\n",
            "88/88 [==============================] - 1s 4ms/step - loss: 0.2450 - binary_accuracy: 0.9093 - val_loss: 0.0913 - val_binary_accuracy: 0.9850\n",
            "Epoch 2/5\n",
            "88/88 [==============================] - 0s 3ms/step - loss: 0.0680 - binary_accuracy: 0.9643 - val_loss: 0.0270 - val_binary_accuracy: 1.0000\n",
            "Epoch 3/5\n",
            "88/88 [==============================] - 0s 5ms/step - loss: 0.0779 - binary_accuracy: 0.9507 - val_loss: 0.0232 - val_binary_accuracy: 0.9967\n",
            "Epoch 4/5\n",
            "88/88 [==============================] - 0s 4ms/step - loss: 0.0765 - binary_accuracy: 0.9407 - val_loss: 0.0598 - val_binary_accuracy: 0.9883\n",
            "Epoch 5/5\n",
            "88/88 [==============================] - 0s 5ms/step - loss: 0.0638 - binary_accuracy: 0.9721 - val_loss: 0.0726 - val_binary_accuracy: 0.9533\n"
          ]
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<keras.callbacks.History at 0x7f0787cb8bd0>"
            ]
          },
          "metadata": {},
          "execution_count": 34
        }
      ],
      "source": [
        "model = keras.Sequential([keras.layers.Dense(1)])\n",
        "model.compile(optimizer=keras.optimizers.RMSprop(learning_rate=0.1),\n",
        "              loss=keras.losses.MeanSquaredError(),\n",
        "              metrics=[keras.metrics.BinaryAccuracy()])\n",
        "\n",
        "indices_permutation = np.random.permutation(len(inputs))\n",
        "shuffled_inputs = inputs[indices_permutation]\n",
        "shuffled_targets = targets[indices_permutation]\n",
        "\n",
        "num_validation_samples = int(0.3 * len(inputs))\n",
        "val_inputs = shuffled_inputs[:num_validation_samples]\n",
        "val_targets = shuffled_targets[:num_validation_samples]\n",
        "training_inputs = shuffled_inputs[num_validation_samples:]\n",
        "training_targets = shuffled_targets[num_validation_samples:]\n",
        "model.fit(\n",
        "    training_inputs,\n",
        "    training_targets,\n",
        "    epochs=5,\n",
        "    batch_size=16,\n",
        "    validation_data=(val_inputs, val_targets)\n",
        ")"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "mKS1lIrmjR1d"
      },
      "source": [
        "### 추론: 훈련한 모델 사용하기"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 35,
      "metadata": {
        "execution": {
          "iopub.execute_input": "2022-05-21T00:58:19.768128Z",
          "iopub.status.busy": "2022-05-21T00:58:19.767816Z",
          "iopub.status.idle": "2022-05-21T00:58:19.874956Z",
          "shell.execute_reply": "2022-05-21T00:58:19.873756Z"
        },
        "id": "O6b3X5hUjR1e",
        "outputId": "c21acbd5-4367-4392-f0f5-6f40e2a25d63",
        "colab": {
          "base_uri": "https://localhost:8080/"
        }
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "5/5 [==============================] - 0s 3ms/step\n",
            "[[0.2212547 ]\n",
            " [0.46757108]\n",
            " [1.3064125 ]\n",
            " [0.36760795]\n",
            " [0.98762196]\n",
            " [1.2589831 ]\n",
            " [0.9364893 ]\n",
            " [0.2468108 ]\n",
            " [0.94096804]\n",
            " [0.4054407 ]]\n"
          ]
        }
      ],
      "source": [
        "predictions = model.predict(val_inputs, batch_size=128)\n",
        "print(predictions[:10])"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "jc1OJ7sJjR1e"
      },
      "source": [
        "## 요약"
      ]
    }
  ],
  "metadata": {
    "colab": {
      "collapsed_sections": [],
      "name": "chapter03_introduction-to-keras-and-tf.i",
      "provenance": []
    },
    "kernelspec": {
      "display_name": "default:Python",
      "language": "python",
      "name": "conda-env-default-py"
    },
    "language_info": {
      "codemirror_mode": {
        "name": "ipython",
        "version": 3
      },
      "file_extension": ".py",
      "mimetype": "text/x-python",
      "name": "python",
      "nbconvert_exporter": "python",
      "pygments_lexer": "ipython3",
      "version": "3.9.10"
    }
  },
  "nbformat": 4,
  "nbformat_minor": 0
}