MLP

README.rst

@@ -257,36 +257,21 @@ Basic Machine Learning
 
 .. _ipythonlinearreg: codes/ipython/basics_in_machine_learning/linearRegressionOneVariable.ipynb
 .. _tutoriallinearreg: https://www.machinelearningmindset.com/linear-regression-with-tensorflow/
-.. _ipythonimageclassification: codes/ipython/basics_in_machine_learning/linearRegressionOneVariable.ipynb
 
 .. |lr| image:: https://colab.research.google.com/assets/colab-badge.svg
   :target: https://colab.research.google.com/github/instillai/TensorFlow-Course/blob/master/codes/ipython/basics_in_machine_learning/linearRegressionOneVariable.ipynb
-.. |ic| image:: https://colab.research.google.com/assets/colab-badge.svg
-  :target: https://colab.research.google.com/github/instillai/TensorFlow-Course/blob/master/codes/ipython/basics_in_machine_learning/ImageClassification.ipynb
+
 
 
 +----+-----------------------------------+--------------------------+-----------------------------------------------+-------------------------------------------+
 | #  |       topic                       |          Run             |  Source Code                                  |  More                                     |
 +====+===================================+==========================+===============================================+===========================================+
 | 1  | Linear Regression                 |       |lr|               | `Notebook <ipythonlinearreg_>`_               | `Tutorial <tutoriallinearreg_>`_          |
-+----+-----------------------------------+--------------------------+-----------------------------------------------+-------------------------------------------+
-| 1  | Image Classification              |       |ic|               | `Notebook <ipythonimageclassification_>`_     |                                           |
 +----+-----------------------------------+--------------------------+-----------------------------------------------+-------------------------------------------+
 
 
- .. _tutoriallinearreg: https://www.machinelearningmindset.com/linear-regression-with-tensorflow/
 
 .. +----+----------------------------+----------------------------------------------------------------------------------------+----------------------------------------------+
-.. | #  |       topic                |   Source Code                                                                          |                                              |
-.. +====+============================+========================================================================================+==============================================+
-.. | 4  | *Linear Models*            |`Linear Regression`_  / `IPython <LinearRegressionipython_>`_                           | `Tutorial      <Documentationlr_>`_          |
-.. +----+----------------------------+----------------------------------------------------------------------------------------+----------------------------------------------+
-.. | 5  | *Predictive Models*        | `Logistic Regression`_  / `IPython <LogisticRegressionipython_>`_                      | `Documentation <LogisticRegDOC_>`_           |
-.. +----+----------------------------+----------------------------------------------------------------------------------------+----------------------------------------------+
-.. | 6  | *Support Vector Machines*  | `Linear SVM`_  / `IPython <LinearSVMipython_>`_                                        |                                              |
-.. +----+----------------------------+----------------------------------------------------------------------------------------+----------------------------------------------+
-.. | 7  | *Support Vector Machines*  |`MultiClass Kernel SVM`_  / `IPython <MultiClassKernelSVMipython_>`_                    |                                              |
-.. +----+----------------------------+----------------------------------------------------------------------------------------+----------------------------------------------+
 
 ==========================
 
@@ -301,15 +286,20 @@ Neural Networks
    :alt: alternate text
    :align: right
 
-+----+-----------------------------------+-----------------------------------------------------------------------------------------------+----------------------------------------------+
-| #  |       topic                       |   Source Code                                                                                 |                                              |
-+====+===================================+===============================================================================================+==============================================+
-| 8  | *Multi Layer Perceptron*          |`Simple Multi Layer Perceptron`_   / `IPython <MultiLayerPerceptronipython_>`_                 |                                              |
-+----+-----------------------------------+-----------------------------------------------------------------------------------------------+----------------------------------------------+
-| 9  | *Convolutional Neural Network*    | `Simple Convolutional Neural Networks`_                                                       |       `Documentation <Documentationcnn_>`_   |
-+----+-----------------------------------+-----------------------------------------------------------------------------------------------+----------------------------------------------+
-| 10 | *Recurrent Neural Network*        | `RNN`_  / `IPython <RNNIpython_>`_                                                            |                                              |
-+----+-----------------------------------+-----------------------------------------------------------------------------------------------+----------------------------------------------+
+
+.. _ipythonmlp: https://github.com/instillai/TensorFlow-Course/blob/master/codes/ipython/neural_networks/mlp.ipynb
+
+
+.. |mlp| image:: https://colab.research.google.com/assets/colab-badge.svg
+ :target: https://colab.research.google.com/github/instillai/TensorFlow-Course/blob/master/codes/ipython/basics_in_machine_learning/ImageClassification.ipynb
+
+
++----+-----------------------------------+--------------------------+-----------------------------------------------+
+| #  |       topic                       |          Run             |  Source Code                                  |
++====+===================================+==========================+===============================================+
+| 1  |  *Multi Layer Perceptron*         |       |mlp|              | `Notebook <ipythonmlp_>`_                     |
++----+-----------------------------------+--------------------------+-----------------------------------------------+
+
 
 .. ~~~~~~~~~~~~
 .. **Welcome**

codes/ipython/neural_networks/mlp.ipynb

@@ -3,7 +3,7 @@
   "nbformat_minor": 0,
   "metadata": {
     "colab": {
-      "name": "ImageClassification.ipynb",
+      "name": "mlp.ipynb",
       "provenance": [],
       "collapsed_sections": []
     },
@@ -16,7 +16,7 @@
     {
       "cell_type": "code",
       "metadata": {
-        "id": "CKfkSFu7nlVg"
+        "id": "x1lvvYoqh_B8"
       },
       "source": [
         "import tensorflow as tf"
@@ -27,8 +27,8 @@
     {
       "cell_type": "code",
       "metadata": {
-        "id": "pznVa656noFn",
-        "outputId": "3fd800b7-db88-4d48-e512-4f7b793a798b",
+        "id": "NljIzKkyiBzz",
+        "outputId": "8e5542ef-854f-4396-a017-2c392702d754",
         "colab": {
           "base_uri": "https://localhost:8080/"
         }
@@ -50,6 +50,8 @@
         {
           "output_type": "stream",
           "text": [
+            "Downloading data from https://storage.googleapis.com/tensorflow/tf-keras-datasets/mnist.npz\n",
+            "11493376/11490434 [==============================] - 0s 0us/step\n",
             "Data type: float64\n",
             "Data type: uint8\n"
           ],
@@ -60,8 +62,8 @@
     {
       "cell_type": "code",
       "metadata": {
-        "id": "zSZB8zVGq46i",
-        "outputId": "884d07cd-ecf3-43b3-b1b0-5440875f4cc8",
+        "id": "gMeQ7B8hiDPp",
+        "outputId": "ef12890e-d3d1-43c2-d32c-dfc92133bde1",
         "colab": {
           "base_uri": "https://localhost:8080/",
           "height": 268
@@ -105,20 +107,11 @@
         }
       ]
     },
-    {
-      "cell_type": "markdown",
-      "metadata": {
-        "id": "6cjdho1KupbS"
-      },
-      "source": [
-        "## Training"
-      ]
-    },
     {
       "cell_type": "code",
       "metadata": {
-        "id": "Z9X1PzYstWKe",
-        "outputId": "635afa44-86dd-495f-b673-5eb4929f0392",
+        "id": "u-_cXLqJiEgq",
+        "outputId": "0d69795b-a262-458f-f5e8-885005740045",
         "colab": {
           "base_uri": "https://localhost:8080/"
         }
@@ -146,25 +139,25 @@
           "output_type": "stream",
           "text": [
             "Epoch 1/10\n",
-            "1875/1875 [==============================] - 4s 2ms/step - loss: 0.2438 - accuracy: 0.9317\n",
+            "1875/1875 [==============================] - 4s 2ms/step - loss: 0.2459 - accuracy: 0.9304\n",
             "Epoch 2/10\n",
-            "1875/1875 [==============================] - 4s 2ms/step - loss: 0.0995 - accuracy: 0.9707\n",
+            "1875/1875 [==============================] - 4s 2ms/step - loss: 0.1022 - accuracy: 0.9690\n",
             "Epoch 3/10\n",
-            "1875/1875 [==============================] - 4s 2ms/step - loss: 0.0692 - accuracy: 0.9793\n",
+            "1875/1875 [==============================] - 4s 2ms/step - loss: 0.0701 - accuracy: 0.9789\n",
             "Epoch 4/10\n",
-            "1875/1875 [==============================] - 4s 2ms/step - loss: 0.0507 - accuracy: 0.9837\n",
+            "1875/1875 [==============================] - 4s 2ms/step - loss: 0.0516 - accuracy: 0.9843\n",
             "Epoch 5/10\n",
-            "1875/1875 [==============================] - 4s 2ms/step - loss: 0.0378 - accuracy: 0.9885\n",
+            "1875/1875 [==============================] - 4s 2ms/step - loss: 0.0389 - accuracy: 0.9880\n",
             "Epoch 6/10\n",
-            "1875/1875 [==============================] - 4s 2ms/step - loss: 0.0284 - accuracy: 0.9910\n",
+            "1875/1875 [==============================] - 4s 2ms/step - loss: 0.0300 - accuracy: 0.9903\n",
             "Epoch 7/10\n",
-            "1875/1875 [==============================] - 4s 2ms/step - loss: 0.0229 - accuracy: 0.9931\n",
+            "1875/1875 [==============================] - 4s 2ms/step - loss: 0.0227 - accuracy: 0.9930\n",
             "Epoch 8/10\n",
-            "1875/1875 [==============================] - 4s 2ms/step - loss: 0.0188 - accuracy: 0.9938\n",
+            "1875/1875 [==============================] - 4s 2ms/step - loss: 0.0191 - accuracy: 0.9941\n",
             "Epoch 9/10\n",
-            "1875/1875 [==============================] - 4s 2ms/step - loss: 0.0149 - accuracy: 0.9953\n",
+            "1875/1875 [==============================] - 4s 2ms/step - loss: 0.0138 - accuracy: 0.9958\n",
             "Epoch 10/10\n",
-            "1875/1875 [==============================] - 4s 2ms/step - loss: 0.0130 - accuracy: 0.9960\n"
+            "1875/1875 [==============================] - 4s 2ms/step - loss: 0.0140 - accuracy: 0.9956\n"
           ],
           "name": "stdout"
         },
@@ -172,7 +165,7 @@
           "output_type": "execute_result",
           "data": {
             "text/plain": [
-              "<tensorflow.python.keras.callbacks.History at 0x7fdf61bfeda0>"
+              "<tensorflow.python.keras.callbacks.History at 0x7f4308281f28>"
             ]
           },
           "metadata": {
@@ -182,20 +175,11 @@
         }
       ]
     },
-    {
-      "cell_type": "markdown",
-      "metadata": {
-        "id": "oD2sc_rNurqN"
-      },
-      "source": [
-        "## Evaluation"
-      ]
-    },
     {
       "cell_type": "code",
       "metadata": {
-        "id": "EfO3wC__tdj5",
-        "outputId": "4d43dbd0-54cf-4904-ae42-ea2adfb97b99",
+        "id": "druwNSZQiGJi",
+        "outputId": "70890649-f7d5-4472-e4af-b7c4a34adf10",
         "colab": {
           "base_uri": "https://localhost:8080/"
         }
@@ -204,13 +188,13 @@
         "eval_loss, eval_acc = model.evaluate(x_test,  y_test, verbose=1)\n",
         "print('Eval accuracy percentage: {:.2f}'.format(eval_acc * 100))\n"
       ],
-      "execution_count": 11,
+      "execution_count": 5,
       "outputs": [
         {
           "output_type": "stream",
           "text": [
-            "313/313 [==============================] - 0s 1ms/step - loss: 0.0781 - accuracy: 0.9793\n",
-            "Eval accuracy percentage: 97.93\n"
+            "313/313 [==============================] - 0s 1ms/step - loss: 0.0882 - accuracy: 0.9781\n",
+            "Eval accuracy percentage: 97.81\n"
           ],
           "name": "stdout"
         }
@@ -219,7 +203,7 @@
     {
       "cell_type": "code",
       "metadata": {
-        "id": "6gf9IHenvlTH"
+        "id": "q5aSajJaiIaM"
       },
       "source": [
         ""

codes/python/neural_networks/cnns.py

+# -*- coding: utf-8 -*-
+"""CNNs.ipynb
+
+Automatically generated by Colaboratory.
+
+Original file is located at
+    https://colab.research.google.com/drive/1ibfKtpxC_hIhZlPbefCoqpAS7jTdyiFw
+"""
+
+import tensorflow as tf
+
+# Load MNIST data
+(x_train, y_train), (x_test, y_test) = tf.keras.datasets.mnist.load_data()
+# Preprocessing
+x_train = x_train / 255.0
+x_test = x_test / 255.0
+
+# Add one domention to make 3D images
+x_train = x_train[...,tf.newaxis]
+x_test = x_test[...,tf.newaxis]
+
+# Track the data type
+dataType, dataShape = x_train.dtype, x_train.shape
+print(f"Data type and shape x_train: {dataType} {dataShape}")
+labelType, labelShape = y_train.dtype, y_train.shape
+print(f"Data type and shape y_train: {labelType} {labelShape}")
+
+im_list = []
+n_samples_to_show = 16
+c = 0
+for i in range(n_samples_to_show):
+  im_list.append(x_train[i])
+# Visualization
+import matplotlib.pyplot as plt
+from mpl_toolkits.axes_grid1 import ImageGrid
+fig = plt.figure(figsize=(4., 4.))
+# Ref: https://matplotlib.org/3.1.1/gallery/axes_grid1/simple_axesgrid.html
+grid = ImageGrid(fig, 111,  # similar to subplot(111)
+                 nrows_ncols=(4, 4),  # creates 2x2 grid of axes
+                 axes_pad=0.1,  # pad between axes in inch.
+                 )
+# Show image grid
+for ax, im in zip(grid, im_list):
+    # Iterating over the grid returns the Axes.
+    ax.imshow(im[:,:,0], 'gray')
+plt.show()
+
+"""## Training"""
+
+# Model building
+NUM_CLASSES = 10
+model = tf.keras.Sequential([
+    tf.keras.layers.Conv2D(16, (3, 3), activation='relu', input_shape=(28, 28, 1)),
+    tf.keras.layers.MaxPooling2D((2, 2)),
+    tf.keras.layers.Conv2D(32, (3, 3), activation='relu'),
+    tf.keras.layers.MaxPooling2D((2, 2)),
+    tf.keras.layers.Conv2D(64, (3, 3), activation='relu'),
+    tf.keras.layers.Flatten(),
+    tf.keras.layers.Dense(32, activation='relu'),
+    tf.keras.layers.Dense(NUM_CLASSES, activation='sigmoid')]
+    )
+
+# Compiling the model with the high-level keras
+model.compile(optimizer='adam',
+              loss=tf.keras.losses.SparseCategoricalCrossentropy(from_logits=False),
+              metrics=['accuracy'])
+
+# Model training
+model.fit(x_train, y_train, epochs=5)
+
+"""## Evaluation"""
+
+eval_loss, eval_acc = model.evaluate(x_test,  y_test, verbose=1)
+print('Eval accuracy percentage: {:.2f}'.format(eval_acc * 100))
+

codes/python/neural_networks/mlp.py

+# -*- coding: utf-8 -*-
+"""mlp.ipynb
+
+Automatically generated by Colaboratory.
+
+Original file is located at
+    https://colab.research.google.com/drive/1SFWk7Ap06ZkvP2HmLhXLiyyqo-ei35M1
+"""
+
+import tensorflow as tf
+
+# Load MNIST data
+(x_train, y_train), (x_test, y_test) = tf.keras.datasets.mnist.load_data()
+# Preprocessing
+x_train = x_train / 255.0
+x_test = x_test / 255.0
+# Track the data type
+dataType = x_train.dtype
+print(f"Data type: {dataType}")
+labelType = y_test.dtype
+print(f"Data type: {labelType}")
+
+im_list = []
+n_samples_to_show = 16
+c = 0
+for i in range(n_samples_to_show):
+  im_list.append(x_train[i])
+# Visualization
+import matplotlib.pyplot as plt
+from mpl_toolkits.axes_grid1 import ImageGrid
+fig = plt.figure(figsize=(4., 4.))
+# Ref: https://matplotlib.org/3.1.1/gallery/axes_grid1/simple_axesgrid.html
+grid = ImageGrid(fig, 111,  # similar to subplot(111)
+                 nrows_ncols=(4, 4),  # creates 2x2 grid of axes
+                 axes_pad=0.1,  # pad between axes in inch.
+                 )
+# Show image grid
+for ax, im in zip(grid, im_list):
+    # Iterating over the grid returns the Axes.
+    ax.imshow(im, 'gray')
+plt.show()
+
+# Model building
+NUM_CLASSES = 10
+model = tf.keras.Sequential([
+    tf.keras.layers.Flatten(input_shape=(28, 28)),
+    tf.keras.layers.Dense(256, activation='relu'),
+    tf.keras.layers.Dense(NUM_CLASSES, activation='sigmoid')
+])
+
+# Compiling the model with the high-level keras
+model.compile(optimizer='adam',
+              loss=tf.keras.losses.SparseCategoricalCrossentropy(from_logits=False),
+              metrics=['accuracy'])
+
+# Model training
+model.fit(x_train, y_train, epochs=10)
+
+eval_loss, eval_acc = model.evaluate(x_test,  y_test, verbose=1)
+print('Eval accuracy percentage: {:.2f}'.format(eval_acc * 100))
+