Chapter 1 - Introduction to Deep Learning

  1. The success of machine learning lies in the right set of features. Feature engineering plays a crucial role in machine learning. If we handcraft the right set of features to predict a certain outcome, then the machine learning algorithms can perform well, but finding and engineering the right set of features is not an easy task. With deep learning, we don't have to handcraft such features. Since deep artificial neural networks (ANNs) employ several layers, they learn the complex intrinsic features and multi-level abstract representation of the data by itself. 

  2. It is basically due to the structure of An ANN. ANNs consist of some n number of layers to perform any computation. We can build an ANN with several layers, where each layer is responsible for learning the intricate patterns in the data. Due to the computational advancements, we can build a network even with hundreds or thousands of layers deep. Since the ANN uses deep layers to perform learning, we call it deep learning, and when an ANN uses deep layers to learn, we call it a deep network. 
  3. The activation function is used to introduce non-linearity to the neural networks. 
  4. When we feed any negative input to the ReLU function, it converts them into zero. The snag for being zero for all negative values is a problem called dying ReLU.
  5. The whole process of moving from the input layer to the output layer to predict output is known as forward propagation. During this propagation, the inputs are multiplied by their respective weights on each layer and an activation function is applied on top of them.
  6. The whole process of backpropagating the network from the output layer to the input layer and updating the weights of the network using gradient descent to minimize the loss is called backpropagation.
  7. Gradient checking is basically used for debugging the gradient descent algorithm and to validate that we have a correct implementation. 
..................Content has been hidden....................

You can't read the all page of ebook, please click here login for view all page.
Reset