Applications of Deep Neural Networks is a free 500 + page book by Jeff Heaton

The contents are as below

The download link is at the bottom of the page

Introduction                                               

 Python Preliminaries

• Assignments                                         
• Your Instructor: Jeff Heaton                                
• Course Resources                                       
• What is Deep Learning                                   
• What is Machine Learning                                  
• Regression                                           
• Classification                                         
• Beyond Classification and Regression                            
• What are Neural Networks                                 
• Why Deep Learning?                                     
• Python for Deep Learning                                  
• Software Installation                                     
• Python Introduction                                     
• Jupyter Notebooks                                      
• Python Versions                                       
• Module Assignment                                     
• Introduction to Python                                  
• Python Lists, Dictionaries, Sets and JSON                       
• Lists and Tuples                                       
• Sets                                              
• Maps/Dictionaries/Hash Tables                               
• More Advanced Lists                                     
• An Introduction to JSON                                  
• File Handling                                       
• Read a CSV File                                       
• Read (stream) a Large CSV File                              
• Read a Text File                                       
• Read an Image                                        
• Functions, Lambdas, and Map/Reduce                         
• Map                                              
• Filter                                             
• Lambda                                            
• Reduce                                            

Python for Machine Learning

• Introduction to Pandas                                  
• Missing Values                                        
• Dealing with Outliers                                    
• Dropping Fields                                       
• Concatenating Rows and Columns                             
• Training and Validation                                   
• Converting a Dataframe to a Matrix                            
• Saving a Dataframe to CSV                                 
• Saving a Dataframe to Pickle                                
• Module Assignment                                     
• Categorical and Continuous Values                           
• Encoding Continuous Values                                
• Encoding Categorical Values as Dummies                         
• Target Encoding for Categoricals                              
• Encoding Categorical Values as Ordinal                          
• Grouping, Sorting, and Shuffling                             
• Shuffling a Dataset                                      
• Sorting a Data Set                                      
• Grouping a Data Set                                     
• Apply and Map                                      
• Using Map with Dataframes                                 
• Using Apply with Dataframes                                
• Feature Engineering with Apply and Map                         
• Feature Engineering                                    
• Calculated Fields                                       
• Google API Keys                                       
• Other Examples: Dealing with Addresses                         

 Introduction to TensorFlow

• Deep Learning and Neural Network Introduction                    
• Classification or Regression                                 
• Neurons and Layers                                     
• Types of Neurons                                       
• Input and Output Neurons                                 
• Hidden Neurons                                       
• Bias Neurons                                         
• Context Neurons                                       
• Other Neuron Types                                     
• Why are Bias Neurons Needed?                               
• Modern Activation Functions                                
• Linear Activation Function                                 
• Rectified Linear Units (ReLU)                               
• Softmax Activation Function                                
• Classic Activation Functions                                
• Step Activation Function                                  
• Sigmoid Activation Function                                
• Hyperbolic Tangent Activation Function                          
• Why ReLU?                                         
• Module Assignment                                     
• Introduction to Tensorflow and Keras                          
• Why TensorFlow                                       
• Deep Learning Tools                                     
• Using TensorFlow Directly                                 
• TensorFlow Linear Algebra Examples                           
• TensorFlow Mandelbrot Set Example                           
• Introduction to Keras                                    
• Simple TensorFlow Regression: MPG                           
• Introduction to Neural Network Hyperparameters                    
• Controlling the Amount of Output                             
• Regression Prediction                                    
• Simple TensorFlow Classification: Iris                           
• Saving and Loading a Keras Neural Network                      
• Early Stopping in Keras to Prevent Overfitting                     
• Early Stopping with Classification                             
• Early Stopping with Regression                               
• Extracting Weights and Manual Network Calculation                 
• Weight Initialization                                     
• Manual Neural Network Calculation                            

 Training for Tabular Data

• Encoding a Feature Vector for Keras Deep Learning                  
• Generate X and Y for a Classification Neural Network                  
• Generate X and Y for a Regression Neural Network                   
• Module Assignment                                     
• Multiclass Classification with ROC and AUC                      
• Binary Classification and ROC Charts                           
• ROC Chart Example                                     
• Multiclass Classification Error Metrics                           
• Calculate Classification Accuracy                              
• Calculate Classification Log Loss                              
• Keras Regression for Deep Neural Networks with RMSE               
• Mean Square Error                                      
• Root Mean Square Error                                  
• Lift Chart                                           
• Training Neural Networks                                 
• Classic Backpropagation                                   
• Momentum Backpropagation                                
• Batch and Online Backpropagation                             
• Stochastic Gradient Descent                                 
• Other Techniques                                       
• ADAM Update                                        
• Methods Compared                                     
• Specifying the Update Rule in Tensorflow                         
• Error Calculation from Scratch                              
• Regression                                           
• Classification                                         

 Regularization and Dropout

• Introduction to Regularization: Ridge and Lasso                    
• L and L Regularization                                  
• Linear Regression                                       
• L (Lasso) Regularization                                  
• L (Ridge) Regularization                                  
• ElasticNet Regularization                                  
• Using K-Fold Cross-validation with Keras                        
• Regression vs Classification K-Fold Cross-Validation                   
• Out-of-Sample Regression Predictions with K-Fold Cross-Validation          
• Classification with Stratified K-Fold Cross-Validation                  
• Training with both a Cross-Validation and a Holdout Set                
• L and L Regularization to Decrease Overfitting                   
• Drop Out for Keras to Decrease Overfitting                       
• Benchmarking Regularization Techniques                        
• Additional Reading on Hyperparameter Tuning                      
• Bootstrapping for Regression                                
• Bootstrapping for Classification                               
• Benchmarking                                        

 Convolutional Neural Networks (CNN) for Computer Vision

• Image Processing in Python                               
• Creating Images (from pixels) in Python                          
• Transform Images in Python (at the pixel level)                     
• Standardize Images                                      
• Adding Noise to an Image                                  
• Module Assignment                                     
• Keras Neural Networks for Digits and Fashion MNIST                
• Computer Vision                                       
• Computer Vision Data Sets                                 
• MNIST Digits Data Set                                   
• MNIST Fashion Data Set                                  
• CIFAR Data Set                                       
• Other Resources                                       
• Convolutional Neural Networks (CNNs)                          
• Convolution Layers                                      
• Max Pooling Layers                                     
• TensorFlow with CNNs                                   
• Access to Data Sets - DIGITS                                
• Display the Digits                                      
• Training/Fitting CNN - DIGITS                              
• Evaluate Accuracy - DIGITS                                
• MNIST Fashion                                        
• Display the Apparel                                     
• Training/Fitting CNN - Fashion                              
• Implementing a ResNet in Keras                             
• Keras Sequence vs Functional Model API                         
• CIFAR Dataset                                        
• ResNet V                                          
• CONTENTS ix
• ResNet V                                          
• Using Your Own Images with Keras                           
• Recognizing Multiple Images with Darknet                       
• How Does DarkNet/YOLO Work?                             
• Using YOLO in Python                                   
• Installing YoloV-TF                                    
• Transfering Weights                                     
• Running DarkFlow (YOLO)                                 
• Module Assignment                                     

 Generative Adversarial Networks

• Introduction to GANS for Image and Data Generation                
• Implementing DCGANs in Keras                            
• Face Generation with StyleGAN and Python                      
• Keras Sequence vs Functional Model API                         
• Generating High Rez GAN Faces with Google CoLab                  
• Run StyleGan From Command Line                           
• Run StyleGAN From Python Code                            
• Examining the Latent Vector                                
• GANS for Semi-Supervised Training in Keras                      
• Semi-Supervised Classification Training                          
• Semi-Supervised Regression Training                            
• Application of Semi-Supervised Regression                        
• An Overview of GAN Research                              
• Select Projects                                        

 Kaggle Data Sets

• Introduction to Kaggle                                  
• Kaggle Ranks                                         
• Typical Kaggle Competition                                 
• How Kaggle Competition Scoring                              
• Preparing a Kaggle Submission                               
• Select Kaggle Competitions                                 
• Module Assignment                                     
• Building Ensembles with Scikit-Learn and Keras                    
• Evaluating Feature Importance                               
• Classification and Input Perturbation Ranking                      
• Regression and Input Perturbation Ranking                        
• Biological Response with Neural Network                         
• What Features/Columns are Important                          
• Neural Network Ensemble                                  
• Architecting Network: Hyperparameters                        
• Number of Hidden Layers and Neuron Counts                       
• Activation Functions                                     
• Advanced Activation Functions                               
• Regularization: L, L, Dropout                              
• Batch Normalization                                     
• Training Parameters                                     
• Experimenting with Hyperparameters                           
• x CONTENTS
• Bayesian Hyperparameter Optimization for Keras                   
• Current Semester’s Kaggle                                
• Iris as a Kaggle Competition                                
• MPG as a Kaggle Competition (Regression)                        
• Module Assignment                                     

 Transfer Learning

• Introduction to Keras Transfer Learning                        
• Transfer Learning Example                                 
• Module Assignment                                     
• Popular Pretrained Neural Networks for Keras                     
• DenseNet                                           
• InceptionResNetV and InceptionV                            
• MobileNet                                           
• MobileNetV                                         
• NASNet                                            
• ResNet, ResNetV, ResNeXt                                
• VGG and VGG                                     
• Xception                                           
• Transfer Learning for Computer Vision and Keras                   
• Transfer                                            
• Transfer Learning for Languages and Keras                       
• Transfer Learning for Keras Feature Engineering                    

 Time Series in Keras

• Time Series Data Encoding                               
• Module Assignment                                     
• Programming LSTM with Keras and TensorFlow                   
• Understanding LSTM                                    
• Simple TensorFlow LSTM Example                            
• Sun Spots Example                                     
• Further Reading for LSTM                                 
• Text Generation with LSTM                              
• Additional Information                                   
• Character-Level Text Generation                              
• Image Captioning with Keras and TensorFlow                    
• Needed Data                                         
• Running Locally                                       
• Clean/Build Dataset From Flickrk                            
• Choosing a Computer Vision Neural Network to Transfer                
• Creating the Training Set                                  
• Using a Data Generator                                   
• Loading Glove Embeddings                                 
• Building the Neural Network                                
• Train the Neural Network                                  
• Generating Captions                                     
• Evaluate Performance on Test Data from Flickerk                    
• Evaluate Performance on My Photos                            
• Module Assignment                                     
• CONTENTS xi
• Temporal CNN in Keras and TensorFlow                       
• Sun Spots Example - CNN                                 

 Natural Language Processing and Speech Recognition

• Getting Started with Spacy in Python                         
• Installing Spacy                                       
• Tokenization                                         
• Sentence Diagramming                                    
• Stop Words                                          
• WordVec and Text Classification                           
• Suggested Software for WordVec                              
• What are Embedding Layers in Keras                         
• Simple Embedding Layer Example                             
• Transferring An Embedding                                 
• Training an Embedding                                   
• Natural Language Processing with Spacy and Keras                 
• Word-Level Text Generation                                
• Learning English from Scratch with Keras and TensorFlow             
• Imports and Utility Functions                                
• Getting the Data                                       
• Building the Vocabulary                                   
• Building the Training and Test Data                            
• Compile the Neural Network                                
• Train the Neural Network                                  
• Evaluate Accuracy                                      
• Adhoc Query                                         

 Reinforcement Learning

• Introduction to the OpenAI Gym                            
• OpenAI Gym Leaderboard                                 
• Looking at Gym Environments                               
• Render OpenAI Gym Environments from CoLab                     
• Introduction to Q-Learning                               
• Introducing the Mountain Car                               
• Programmed Car                                       
• Reinforcement Learning                                   
• Running and Observing the Agent                             
• Inspecting the Q-Table                                    
• Keras Q-Learning in the OpenAI Gym                         
• DQN and the Cart-Pole Problem                              
• Hyperparameters                                       
• Environment                                         
• Agent                                             
• Policies                                            
• Metrics and Evaluation                                   
• Replay Buffer                                         
• Data Collection                                        
• Training the agent                                      
• Visualization                                         
• KS-Statistic                                          
• Detecting Drift between Training and Testing Datasets by Training          
• Using a Keras Deep Neural Network with a Web Application            
• Converting Keras to CoreML                                
• Creating an IOS CoreML Application                           
• More Reading                                         
• Other Neural Network Techniques
• What is AutoML                                     
• AutoML from your Local Computer                            
• AutoML from Google Cloud                                 
• A Simple AutoML System                                  
• Running My Sample AutoML Program                          
• Using Denoising AutoEncoders in Keras                        
• Function Approximation                                   
• Multi-Output Regression                                  
• Simple Autoencoder                                     
• Autoencode (single image)                                  
• Standardize Images                                      
• Image Autoencoder (multi-image)                             
• Adding Noise to an Image                                  
• Denoising Autoencoder                                   
• Anomaly Detection in Keras                              
• Read in KDD Data Set                                  
• Preprocessing                                         
• Training the Autoencoder                                  
• Detecting an Anomaly                                    
• Training an Intrusion Detection System with KDD                 
• Read in Raw KDD- Dataset                               
• Analyzing a Dataset                                     
• Encode the feature vector                                  
• Train the Neural Network                                  
• New Technologies                                     
• Neural Structured Learning (NSL)                             
• Bert, AlBert, and Other NLP Technologies                        
• Explainability Frameworks                               

Advanced/Other Topics

• Flask and Deep Learning Web Services                        
• Flask Hello World                                      
• MPG Flask                                          
• Flask MPG Client                                      
• Images and Web Services                                  
• Part : Deploying a Model to AWS                               
• Train Model (optionally, outside of AWS)                         
• Next Step: Convert Model (must use AWS SageMaker Notebook)           
• Step Set up                                         
• Step Load the Keras model using the JSON and weights file              
• Step Export the Keras model to the TensorFlow ProtoBuf format (must use AWS
• SageMaker Notebook)                                    
• Step Convert TensorFlow model to a SageMaker readable format (must use AWS
• SageMaker Notebook)                                    
• Tar the entire directory and upload to S                         
• Step Deploy the trained model (must use AWS SageMaker Notebook)         
• Test Model Deployment (optionally, outside of AWS)              
• Call the end point                                      
• Additional Reading                                      
• Using a Keras Deep Neural Network with a Web Application            
• When to Retrain Your Neural Network                        
• Preprocessing the Sberbank Russian Housing Market Data               
• KS-Statistic                                          
• Detecting Drift between Training and Testing Datasets by Training          
• Using a Keras Deep Neural Network with a Web Application            
• Converting Keras to CoreML                                
• Creating an IOS CoreML Application                           
• More Reading                                         
• Other Neural Network Techniques
• What is AutoML                                     
• AutoML from your Local Computer                            
• AutoML from Google Cloud                                 
• A Simple AutoML System                                  
• Running My Sample AutoML Program                          
• Using Denoising AutoEncoders in Keras                        
• Function Approximation                                   
• Multi-Output Regression                                  
• Simple Autoencoder                                     
• Autoencode (single image)                                  
• Standardize Images                                      
• Image Autoencoder (multi-image)                             
• Adding Noise to an Image                                  
• Denoising Autoencoder                                   
• Anomaly Detection in Keras                              
• Read in KDD Data Set                                  
• Preprocessing                                         
• Training the Autoencoder                                  
• Detecting an Anomaly                                    
• Training an Intrusion Detection System with KDD                 
• Read in Raw KDD- Dataset                               
• Analyzing a Dataset                                     
• Encode the feature vector                                  
• Train the Neural Network                                  
• New Technologies                                     
• Neural Structured Learning (NSL)                             
• Bert, AlBert, and Other NLP Technologies                        
• Explainability Frameworks                                 

Book download link HERE

Image source pixabay