Table of contents : Cover Table of Contents Series Page Title Page Copyright Page Preface 1 Introduction to Deep Learning 1.1 History of Deep Learning 1.2 A Probabilistic Theory of Deep Learning 1.3 Back Propagation and Regularization 1.4 Batch Normalization and VC Dimension 1.5 Neural Nets—Deep and Shallow Networks 1.6 Supervised and Semi-Supervised Learning 1.7 Deep Learning and Reinforcement Learning References 2 Basics of TensorFlow 2.1 Tensors 2.2 Computational Graph and Session 2.3 Constants, Placeholders, and Variables 2.4 Creating Tensor 2.5 Working on Matrices 2.6 Activation Functions 2.7 Loss Functions 2.8 Common Loss Function 2.9 Optimizers 2.10 Metrics References 3 Understanding and Working with Keras 3.1 Major Steps to Deep Learning Models 3.2 Load Data 3.3 Pre-Process Data 3.4 Define the Model 3.5 Compile the Model 3.6 Fit and Evaluate the Mode 3.7 Prediction 3.8 Save and Reload the Model 3.9 Additional Steps to Improve Keras Models 3.10 Keras with TensorFlow References 4 Multilayer Perceptron 4.1 Artificial Neural Network 4.2 Single-Layer Perceptron 4.3 Multilayer Perceptron 4.4 Logistic Regression Model 4.5 Regression to MLP in TensorFlow 4.6 TensorFlow Steps to Build Models 4.7 Linear Regression in TensorFlow 4.8 Logistic Regression Mode in TensorFlow 4.9 Multilayer Perceptron in TensorFlow 4.10 Regression to MLP in Keras 4.11 Log-Linear Model 4.12 Keras Neural Network for Linear Regression 4.13 Keras Neural Network for Logistic Regression 4.14 MLPs on the Iris Data 4.15 MLPs on MNIST Data (Digit Classification) 4.16 MLPs on Randomly Generated Data References 5 Convolutional Neural Networks in Tensorflow 5.1 CNN Architectures 5.2 Properties of CNN Representations 5.3 Convolution Layers, Pooling Layers – Strides - Padding and Fully Connected Layer 5.4 Why TensorFlow for CNN Models? 5.5 TensorFlow Code for Building an Image Classifier for MNIST Data 5.6 Using a High-Level API for Building CNN Models 5.7 CNN in Keras 5.8 Building an Image Classifier for MNIST Data in Keras 5.9 Building an Image Classifier with CIFAR-10 Data 5.10 Define the Model Architecture 5.11 Pre-Trained Models References 6 RNN and LSTM 6.1 Concept of RNN 6.2 Concept of LSTM 6.3 Modes of LSTM 6.4 Sequence Prediction 6.5 Time-Series Forecasting with the LSTM Model 6.6 Speech to Text 6.7 Examples Using Each API 6.8 Text-to-Speech Conversion 6.9 Cognitive Service Providers 6.10 The Future of Speech Analytics References 7 Developing Chatbot’s Face Detection and Recognition 7.1 Why Chatbots? 7.2 Designs and Functions of Chatbot’s 7.3 Steps for Building a Chatbot’s 7.4 Best Practices of Chatbot Development 7.5 Face Detection 7.6 Face Recognition 7.7 Face Analysis 7.8 OpenCV—Detecting a Face, Recognition and Face Analysis 7.9 Deep Learning–Based Face Recognition 7.10 Transfer Learning 7.11 API’s References 8 Advanced Deep Learning 8.1 Deep Convolutional Neural Networks (AlexNet) 8.2 Networks Using Blocks (VGG) 8.3 Network in Network (NiN) 8.4 Networks with Parallel Concatenations (GoogLeNet) 8.5 Residual Networks (ResNet) 8.6 Densely Connected Networks (DenseNet) 8.7 Gated Recurrent Units (GRU) 8.8 Long Short-Term Memory (LSTM) 8.9 Deep Recurrent Neural Networks (D-RNN) 8.10 Bidirectional Recurrent Neural Networks (Bi-RNN) 8.11 Machine Translation and the Dataset 8.12 Sequence to Sequence Learning References 9 Enhanced Convolutional Neural Network 9.1 Introduction 9.2 Deep Learning-Based Architecture for Absence Seizure Detection 9.3 EEG Signal Pre-Processing Strategy and Channel Selection 9.4 Input Formulation and Augmentation of EEG Signal for Deep Learning Model 9.5 Deep Learning Based Feature Extraction and Classification 9.6 Performance Analysis 9.7 Summary References 10 Conclusion 10.1 Introduction 10.2 Future Research Direction and Prospects 10.3 Research Challenges in Deep Learning 10.4 Practical Deep Learning Case Studies 10.5 Summary References Index End User License Agreement