• VIEW CHAPTER DETAILS

      Online Courses
      Important Links
      Digital Books
      Videos and Tutorials

    • VIEW CHAPTER DETAILS


      What is Keras

      • What is Keras?, 
      • Theano vs Keras, 
      • Keras, Why use Keras?,
      •  Keras + TensorFlow, 
      • Feature Engineering, 
      • Unstructured data,
      • So, when to use neural networks?

      Your first neural network

      • A neural network?, 
      • Parameters, Gradient descent.
      •  The sequential API. 
      • Defining a neural network.
      •  Adding activations. 
      • Summarize your model!, 
      • Visualize parameter

      Surviving a meteor strike

      • Surviving a meteor strike,
      • Recap, Compiling,
      • Training,
      • Predicting,
      • Evaluating,
      • The problem at hand, Scientific prediction,
      • Your task

    • VIEW CHAPTER DETAILS

                                 

       Binary classification

      • When to use binary classification?
      • Our dataset
      • Pair plots
      • The NN architecture
      • The sigmoid function
      • Compiling, training, predicting
      • Results

      Multi-class classification

      • Throwing darts
      • The dataset
      • The architecture
      • The output layer
      • Multi-class model
      • Categorical cross-entropy
      • Preparing a dataset
      • One-hot encoding

      Multi-label classification

      • Real-world examples
      • Multi-class vs multi-label
      • The architecture
      • Sigmoid outputs
      • Compile and train
      • An advantage
      • An irrigation machine

      Keras callbacks

      • What is a callback?
      • Callbacks in Keras
      • A callback you've been missing
      • History plots
      • Early stopping
      • Model checkpoint

    • VIEW CHAPTER DETAILS

          Learning curves

      • Learning curves
      • Loss curve
      • Accuracy curve
      • Overfitting
      • Unstable curves
      • Can we benefit from more data
      • Coding train size comparison

      Activation functions

      • Activation zoo
      • Effects of activation functions
      • Which activation function to use?
      • Comparing activation functions

      Batch size and batch normalization

      • Batches
      • Mini-batch
      • Effects of batch sizes
      • Batch size in Keras
      • Normalization in machine learning
      • Reasons for batch normalization
      • Batch normalization advantages
      • Batch normalization in Keras

      Hyperparameter tuning

      • Neural network hyperparameters
      • Sklearn recap
      • Turn a Keras model into a Sklearn estimator
      • Cross-validation
      • Tips for neural networks hyperparameter tuning
      • A random search on Keras models
      • Tuning other hyperparameters

       

    • VIEW CHAPTER DETAILS

      Tensors, layers, and autoencoders

      • Accessing Keras layers
      • What are tensors?
      • Keras backend
      • Introducing autoencoders
      • Autoencoder use cases
      • Building a simple autoencoder
      • Breaking it into an encoder

      Intro to CNN's

      • How do they work?
      • Convolutions demonstration
      • Typical architectures
      • Input shape to convolutional neural networks
      • How to build a simple convolutional net in Keras?
      • Deep convolutional models
      • Pre-processing images for ResNet50
      • Using the ResNet50 model in Keras
      • What is going on inside a convnet?

      Intro to LSTMs

      • What are RNNs?
      • What are LSTMs?
      • What are LSTMs?
      • When to use LSTMs?
      • LSTMs + Text
      • Embeddings
      • Sequence preparation
      • Building a LSTM model
      • Building a LSTM model