Merge pull request #727 from bestfw/LSTM-assignment

Commit recovered changes

open-machine-learning-jupyter-book/_toc.yml

@@ -213,6 +213,7 @@ parts:
     - file: assignments/deep-learning/cnn/how-to-choose-cnn-architecture-mnist
     - file: assignments/deep-learning/cnn/sign-language-digits-classification-with-cnn
     - file: assignments/deep-learning/cnn/object-recognition-in-images-using-cnn
+    - file: assignments/deep-learning/cnn/image-classification
     - file: assignments/deep-learning/tensorflow/intro_to_tensorflow_for_deeplearning
     - file: assignments/deep-learning/lstm/bitcoin-lstm-model-with-tweet-volume-and-sentiment
     - file: assignments/deep-learning/rnn/google-stock-price-prediction-rnn
@@ -225,6 +226,7 @@ parts:
     - file: assignments/deep-learning/dqn/dqn-on-foreign-exchange-market
     - file: assignments/deep-learning/gan/art-by-gan
     - file: assignments/deep-learning/gan/gan-introduction
+    - file: assignments/deep-learning/image-segmentation/comparing-edge-based-and-region-based-segmentation
     - file: assignments/deep-learning/difussion-model/denoising-difussion-model
     - file: assignments/deep-learning/object-detection/car-object-detection
     - file: assignments/deep-learning/overview/basic-classification-classify-images-of-clothing

open-machine-learning-jupyter-book/deep-learning/image-classification.ipynb

@@ -219,8 +219,8 @@
    "source": [
     "- Using small 3×3 filters to replace large convolutional kernels.\n",
     "- After replacing the convolution kernel, the convolution layers have the same perceptual field. \n",
-    "- Each layer is trained by Re LU activation function and batch gradient descent after convolution operation.\n",
-    "- It is verified that increasing the network depth can improve the model performance Although, VGG has achieved good results in image classification and localization problems in 2014 due to its deeper network structure and low computational complexity, it uses 140 million parameters and is computationally intensive, which is its shortcoming."
+    "- Each layer is trained by ReLU activation function and batch gradient descent after convolution operation.\n",
+    "- It is verified that increasing the network depth can improve the model performance. Although, VGG has achieved good results in image classification and localization problems in 2014 due to its deeper network structure and low computational complexity, it uses 140 million parameters and is computationally intensive, which is its shortcoming."
    ]
   },
   {
@@ -1156,7 +1156,7 @@
    "id": "b4552758",
    "metadata": {},
    "source": [
-    "TBD."
+    "Assignment - [Image classification](../../assignments/deep-learning/cnn/image-classification.ipynb)"
    ]
   },
   {

open-machine-learning-jupyter-book/deep-learning/image-segmentation.ipynb

@@ -2045,7 +2045,7 @@
    "id": "a0994ed1",
    "metadata": {},
    "source": [
-    "TBD."
+    "Assignment - [Comparing edge-based and region-based segmentation](../../assignments/deep-learning/image-segmentation/comparing-edge-based-and-region-based-segmentation.ipynb)"
    ]
   },
   {

open-machine-learning-jupyter-book/deep-learning/lstm.ipynb

@@ -4770,7 +4770,7 @@
    "source": [
     "## Your turn! 🚀\n",
     "\n",
-    "Practice the Long-Short Term Memory Networks by following this TBD.\n",
+    "Assignment - [Bitcoin lstm model with tweet volume and sentiment](../../assignments/deep-learning/lstm/bitcoin-lstm-model-with-tweet-volume-and-sentiment.ipynb)\n",
     "\n",
     "## Acknowledgments\n",
     "\n",