St. Xavier's University Library Search for 'Provider:Shroff publishers & Distributors '

Digital transformation game plan : 34 tenets for masterfully merging technology and business

]]> By O'Brien, Gary.; Guo, Xiao.; Mason, Mike..
Mumbai Shroff Publishers & Distributors 2019 .
344p. 9789352139309

]]> Place hold on Digital transformation game plan

Hands-on machine learning with scikit-learn, keras, and tensorflow : concepts, tools, and techniques to build intelligent systems

]]> By Geron, Aurelien .
New Delhi Shroff publishers & distributors 2019 .
xxv, 819 , Part I. The Fundamentals of Machine Learning 1. The Machine Learning Landscape. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 What Is Machine Learning? 2 Why Use Machine Learning? 2 Examples of Applications 5 Types of Machine Learning Systems 7 Supervised/Unsupervised Learning 7 Batch and Online Learning 14 Instance-Based Versus Model-Based Learning 17 Main Challenges of Machine Learning 23 Insufficient Quantity of Training Data 23 Nonrepresentative Training Data 25 Poor-Quality Data 26 Irrelevant Features 27 Overfitting the Training Data 27 Underfitting the Training Data 29 Stepping Back 30 Testing and Validating 30 Hyperparameter Tuning and Model Selection 31 Data Mismatch 32 Exercises 33 2. End-to-End Machine Learning Project. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35 Working with Real Data 35 iii Look at the Big Picture 37 Frame the Problem 37 Select a Performance Measure 39 Check the Assumptions 42 Get the Data 42 Create the Workspace 42 Download the Data 46 Take a Quick Look at the Data Structure 47 Create a Test Set 51 Discover and Visualize the Data to Gain Insights 56 Visualizing Geographical Data 56 Looking for Correlations 58 Experimenting with Attribute Combinations 61 Prepare the Data for Machine Learning Algorithms 62 Data Cleaning 63 Handling Text and Categorical Attributes 65 Custom Transformers 68 Feature Scaling 69 Transformation Pipelines 70 Select and Train a Model 72 Training and Evaluating on the Training Set 72 Better Evaluation Using Cross-Validation 73 Fine-Tune Your Model 75 Grid Search 76 Randomized Search 78 Ensemble Methods 78 Analyze the Best Models and Their Errors 78 Evaluate Your System on the Test Set 79 Launch, Monitor, and Maintain Your System 80 Try It Out! 83 Exercises 84 3. Classification. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 85 MNIST 85 Training a Binary Classifier 88 Performance Measures 88 Measuring Accuracy Using Cross-Validation 89 Confusion Matrix 90 Precision and Recall 92 Precision/Recall Trade-off 93 The ROC Curve 97 Multiclass Classification 100 iv | Table of Contents Error Analysis 102 Multilabel Classification 106 Multioutput Classification 107 Exercises 108 4. Training Models. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 111 Linear Regression 112 The Normal Equation 114 Computational Complexity 117 Gradient Descent 118 Batch Gradient Descent 121 Stochastic Gradient Descent 124 Mini-batch Gradient Descent 127 Polynomial Regression 128 Learning Curves 130 Regularized Linear Models 134 Ridge Regression 135 Lasso Regression 137 Elastic Net 140 Early Stopping 141 Logistic Regression 142 Estimating Probabilities 143 Training and Cost Function 144 Decision Boundaries 145 Softmax Regression 148 Exercises 151 5. Support Vector Machines. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 153 Linear SVM Classification 153 Soft Margin Classification 154 Nonlinear SVM Classification 157 Polynomial Kernel 158 Similarity Features 159 Gaussian RBF Kernel 160 Computational Complexity 162 SVM Regression 162 Under the Hood 164 Decision Function and Predictions 165 Training Objective 166 Quadratic Programming 167 The Dual Problem 168 Kernelized SVMs 169 Table of Contents | v Online SVMs 172 Exercises 174 6. Decision Trees. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 175 Training and Visualizing a Decision Tree 175 Making Predictions 176 Estimating Class Probabilities 178 The CART Training Algorithm 179 Computational Complexity 180 Gini Impurity or Entropy? 180 Regularization Hyperparameters 181 Regression 183 Instability 185 Exercises 186 7. Ensemble Learning and Random Forests. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 189 Voting Classifiers 189 Bagging and Pasting 192 Bagging and Pasting in Scikit-Learn 194 Out-of-Bag Evaluation 195 Random Patches and Random Subspaces 196 Random Forests 197 Extra-Trees 198 Feature Importance 198 Boosting 199 AdaBoost 200 Gradient Boosting 203 Stacking 208 Exercises 211 8. Dimensionality Reduction. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 213 The Curse of Dimensionality 214 Main Approaches for Dimensionality Reduction 215 Projection 215 Manifold Learning 218 PCA 219 Preserving the Variance 219 Principal Components 220 Projecting Down to d Dimensions 221 Using Scikit-Learn 222 Explained Variance Ratio 222 Choosing the Right Number of Dimensions 223 vi | Table of Contents PCA for Compression 224 Randomized PCA 225 Incremental PCA 225 Kernel PCA 226 Selecting a Kernel and Tuning Hyperparameters 227 LLE 230 Other Dimensionality Reduction Techniques 232 Exercises 233 9. Unsupervised Learning Techniques. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 235 Clustering 236 K-Means 238 Limits of K-Means 248 Using Clustering for Image Segmentation 249 Using Clustering for Preprocessing 251 Using Clustering for Semi-Supervised Learning 253 DBSCAN 255 Other Clustering Algorithms 258 Gaussian Mixtures 260 Anomaly Detection Using Gaussian Mixtures 266 Selecting the Number of Clusters 267 Bayesian Gaussian Mixture Models 270 Other Algorithms for Anomaly and Novelty Detection 274 Exercises 275 Part II. Neural Networks and Deep Learning 10. Introduction to Artificial Neural Networks with Keras. . . . . . . . . . . . . . . . . . . . . . . . . . 279 From Biological to Artificial Neurons 280 Biological Neurons 281 Logical Computations with Neurons 283 The Perceptron 284 The Multilayer Perceptron and Backpropagation 289 Regression MLPs 292 Classification MLPs 294 Implementing MLPs with Keras 295 Installing TensorFlow 2 296 Building an Image Classifier Using the Sequential API 297 Building a Regression MLP Using the Sequential API 307 Building Complex Models Using the Functional API 308 Using the Subclassing API to Build Dynamic Models 313 Table of Contents | vii Saving and Restoring a Model 314 Using Callbacks 315 Using TensorBoard for Visualization 317 Fine-Tuning Neural Network Hyperparameters 320 Number of Hidden Layers 323 Number of Neurons per Hidden Layer 325 Learning Rate, Batch Size, and Other Hyperparameters 325 Exercises 327 11. Training Deep Neural Networks. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 331 The Vanishing/Exploding Gradients Problems 332 Glorot and He Initialization 333 Nonsaturating Activation Functions 335 Batch Normalization 338 Gradient Clipping 345 Reusing Pretrained Layers 345 Transfer Learning with Keras 347 Unsupervised Pretraining 349 Pretraining on an Auxiliary Task 350 Faster Optimizers 351 Momentum Optimization 351 Nesterov Accelerated Gradient 353 AdaGrad 354 RMSProp 355 Adam and Nadam Optimization 356 Learning Rate Scheduling 359 Avoiding Overfitting Through Regularization 364 ℓ1 and ℓ2 Regularization 364 Dropout 365 Monte Carlo (MC) Dropout 368 Max-Norm Regularization 370 Summary and Practical Guidelines 371 Exercises 373 12. Custom Models and Training with TensorFlow. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 375 A Quick Tour of TensorFlow 376 Using TensorFlow like NumPy 379 Tensors and Operations 379 Tensors and NumPy 381 Type Conversions 381 Variables 382 Other Data Structures 383 viii | Table of Contents Customizing Models and Training Algorithms 384 Custom Loss Functions 384 Saving and Loading Models That Contain Custom Components 385 Custom Activation Functions, Initializers, Regularizers, and Constraints 387 Custom Metrics 388 Custom Layers 391 Custom Models 394 Losses and Metrics Based on Model Internals 397 Computing Gradients Using Autodiff 399 Custom Training Loops 402 TensorFlow Functions and Graphs 405 AutoGraph and Tracing 407 TF Function Rules 409 Exercises 410 13. Loading and Preprocessing Data with TensorFlow. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 413 The Data API 414 Chaining Transformations 415 Shuffling the Data 416 Preprocessing the Data 419 Putting Everything Together 420 Prefetching 421 Using the Dataset with tf.keras 423 The TFRecord Format 424 Compressed TFRecord Files 425 A Brief Introduction to Protocol Buffers 425 TensorFlow Protobufs 427 Loading and Parsing Examples 428 Handling Lists of Lists Using the SequenceExample Protobuf 429 Preprocessing the Input Features 430 Encoding Categorical Features Using One-Hot Vectors 431 Encoding Categorical Features Using Embeddings 433 Keras Preprocessing Layers 437 TF Transform 439 The TensorFlow Datasets (TFDS) Project 441 Exercises 442 14. Deep Computer Vision Using Convolutional Neural Networks. . . . . . . . . . . . . . . . . . . 445 The Architecture of the Visual Cortex 446 Convolutional Layers 448 Filters 450 Stacking Multiple Feature Maps 451 Table of Contents | ix TensorFlow Implementation 453 Memory Requirements 456 Pooling Layers 456 TensorFlow Implementation 458 CNN Architectures 460 LeNet-5 463 AlexNet 464 GoogLeNet 466 VGGNet 470 ResNet 471 Xception 474 SENet 476 Implementing a ResNet-34 CNN Using Keras 478 Using Pretrained Models from Keras 479 Pretrained Models for Transfer Learning 481 Classification and Localization 483 Object Detection 485 Fully Convolutional Networks 487 You Only Look Once (YOLO) 489 Semantic Segmentation 492 Exercises 496 15. Processing Sequences Using RNNs and CNNs. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 497 Recurrent Neurons and Layers 498 Memory Cells 500 Input and Output Sequences 501 Training RNNs 502 Forecasting a Time Series 503 Baseline Metrics 505 Implementing a Simple RNN 505 Deep RNNs 506 Forecasting Several Time Steps Ahead 508 Handling Long Sequences 511 Fighting the Unstable Gradients Problem 512 Tackling the Short-Term Memory Problem 514 Exercises 523 16. Natural Language Processing with RNNs and Attention. . . . . . . . . . . . . . . . . . . . . . . . 525 Generating Shakespearean Text Using a Character RNN 526 Creating the Training Dataset 527 How to Split a Sequential Dataset 527 Chopping the Sequential Dataset into Multiple Windows 528 x | Table of Contents Building and Training the Char-RNN Model 530 Using the Char-RNN Model 531 Generating Fake Shakespearean Text 531 Stateful RNN 532 Sentiment Analysis 534 Masking 538 Reusing Pretrained Embeddings 540 An Encoder–Decoder Network for Neural Machine Translation 542 Bidirectional RNNs 546 Beam Search 547 Attention Mechanisms 549 Visual Attention 552 Attention Is All You Need: The Transformer Architecture 554 Recent Innovations in Language Models 563 Exercises 565 17. Representation Learning and Generative Learning Using Autoencoders and GANs. 567 Efficient Data Representations 569 Performing PCA with an Undercomplete Linear Autoencoder 570 Stacked Autoencoders 572 Implementing a Stacked Autoencoder Using Keras 572 Visualizing the Reconstructions 574 Visualizing the Fashion MNIST Dataset 574 Unsupervised Pretraining Using Stacked Autoencoders 576 Tying Weights 577 Training One Autoencoder at a Time 578 Convolutional Autoencoders 579 Recurrent Autoencoders 580 Denoising Autoencoders 581 Sparse Autoencoders 582 Variational Autoencoders 586 Generating Fashion MNIST Images 590 Generative Adversarial Networks 592 The Difficulties of Training GANs 596 Deep Convolutional GANs 598 Progressive Growing of GANs 601 StyleGANs 604 Exercises 607 18. Reinforcement Learning. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 609 Learning to Optimize Rewards 610 Policy Search 612 Table of Contents | xi Introduction to OpenAI Gym 613 Neural Network Policies 617 Evaluating Actions: The Credit Assignment Problem 619 Policy Gradients 620 Markov Decision Processes 625 Temporal Difference Learning 629 Q-Learning 630 Exploration Policies 632 Approximate Q-Learning and Deep Q-Learning 633 Implementing Deep Q-Learning 634 Deep Q-Learning Variants 639 Fixed Q-Value Targets 639 Double DQN 640 Prioritized Experience Replay 640 Dueling DQN 641 The TF-Agents Library 642 Installing TF-Agents 643 TF-Agents Environments 643 Environment Specifications 644 Environment Wrappers and Atari Preprocessing 645 Training Architecture 649 Creating the Deep Q-Network 650 Creating the DQN Agent 652 Creating the Replay Buffer and the Corresponding Observer 654 Creating Training Metrics 655 Creating the Collect Driver 656 Creating the Dataset 658 Creating the Training Loop 661 Overview of Some Popular RL Algorithms 662 Exercises 664 19. Training and Deploying TensorFlow Models at Scale. . . . . . . . . . . . . . . . . . . . . . . . . . . 667 Serving a TensorFlow Model 668 Using TensorFlow Serving 668 Creating a Prediction Service on GCP AI Platform 677 Using the Prediction Service 682 Deploying a Model to a Mobile or Embedded Device 685 Using GPUs to Speed Up Computations 689 Getting Your Own GPU 690 Using a GPU-Equipped Virtual Machine 692 Colaboratory 693 Managing the GPU RAM 694 xii | Table of Contents Placing Operations and Variables on Devices 697 Parallel Execution Across Multiple Devices 699 Training Models Across Multiple Devices 701 Model Parallelism 701 Data Parallelism 704 Training at Scale Using the Distribution Strategies API 709 Training a Model on a TensorFlow Cluster 711 Running Large Training Jobs on Google Cloud AI Platform 714 Black Box Hyperparameter Tuning on AI Platform 716 Exercises 717 Thank You! 718 A. Exercise Solutions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 719 B. Machine Learning Project Checklist. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 755 C. SVM Dual Problem. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 761 D. Autodiff. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 765 E. Other Popular ANN Architectures. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 773 F. Special Data Structures. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 783 G. TensorFlow Graphs. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 791 Index. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 801 9789352139057

]]> Place hold on Hands-on machine learning with scikit-learn, keras, and tensorflow

Make : Bluetooth LE projects for Arduino, Raspberry Pi, and Smartphones

]]> By Allan, Alasdair .
New Delhi Shroff publishers & distributors 2016 .
xv, 238 9789352132997

]]> Place hold on Make

Enterprise IoT

]]> Kolkata Shroff publishers & distributors 2016 .
xvii, 464 , Table of Contents Foreword. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xi Preface. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xiii 1. Overture. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 Mission: 50 Billion Connected Devices by 2020 1 Customer Perspective: Value-Added Services 3 Manufacturer Perspective: Connected Asset Lifecycle Management 3 Servitization: The Next Logical Step? 5 Prerequisite: Operator Approach 7 Impact: Disruption Versus Evolution 8 Clash of Two Worlds: Machine Camp Versus Internet Camp 9 Difficulty of Finding the Right Service 11 Foundation: Digitization of the Physical World 13 Critical: Security and Data Privacy 14 Timing: Why Now? 14 Keynote Contribution: IoT and Smart, Connected Products 16 2. Enterprise IoT. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23 From M2M Toward the IoT 23 Subnets of Things 26 Focus of this Book 27 Domain Focus 28 Definitions of Key Terms in IoT 29 v Part I. IoT Application Domains and Case Studies 3. Smart Energy. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39 Influence of Digitization 41 Generation 43 Transmission 44 Distribution and Metering 44 Storage 45 Marketing, Sales, and Service 45 Customers 46 When Is All This Going to Happen? 46 Conclusions 47 Energy Case Studies 49 Smart Monitoring and Diagnostics Systems at Major Power Plants 49 Asset Integration Architecture of Smart M&D 54 Lessons Learned 54 Microgrids and Virtual Power Plants 56 VPP/MMS: Functional Overview 57 Case Study: Smart City Rheintal 59 Smart Energy in the Chemical Industry 63 4. Manufacturing and Industry. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 67 Integrated Production for Integrated Products 67 Sales/Marketing and New Business Models 70 End-to-End Digital Engineering 71 Manufacturing 72 IoT Service Implementation 74 IoT Service Operations 74 Aftermarket Services 75 Work Environment 76 Adaptive Logistics and Value-Added Networks 76 Other Industrial Applications 77 Industry Initiatives 77 Industry 4.0 77 Industrial Internet Consortium 80 Case Studies: Overview 81 Case Study: Smart Factory 81 Asset Integration Architecture 84 Conclusions and Outlook 86 Case Study: Intelligent Lot Tracking 87 Technical Architecture 89 Conclusions and Outlook 90 Case Study: Cleaning Service Industry and Technology 92 Kärcher Fleet Management Solution 92 Portal 94 vi TABLE OF CONTENTS Asset Integration Architecture 95 Lessons Learned 97 Case Study: Global Cold Chain Management 99 Functional Solution Overview 101 Technical Solution Details and AIA 104 Lessons Learned and Recommendations 106 Case Study: LHCb Experiment at CERN 107 LHCb and Data Management 108 LHCb and Physical Data Analysis 111 5. Connected Vehicle. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 117 Car Dashboard and Infotainment 119 Value-Added Services 121 eCall 122 bCall 124 Stolen Vehicle Recovery 124 Usage-Based Insurance 125 So, Why No Open Car App Platform (Yet)? 128 Connected Enterprise Solutions 129 Fleet Management 129 Systematic Field Data 132 eMobility 134 EV Charging Services 134 eRoaming 137 EV Remote Management 138 EVs and Cross-Energy Management 140 Car Sharing 140 Intermodal Services 142 Vehicle Functions (Toward Automated Driving) 142 The Roadmap Toward Automated Driving 142 Automated Driving: Technologies 145 Automated Driving: System Architecture 147 Digital Horizon 149 Parking 150 Outlook 152 6. Smart City. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 155 Key Drivers 155 Smart City Examples 156 Smart City Projects in Chicago: Implementing Live Pilot Tests 156 Smart City Projects in Rio de Janeiro: Using Safety to Attract New Business 157 Smart City Projects in Stockholm: Structuring a Dialog with Citizens and Private Companies 157 Smart City Projects in Boston: Connecting Citizens with City Services 158 Smart City Projects in Hong Kong: Focusing in on ICT 160 TABLE OF CONTENTS vii Business Models and KPIs for Smart Cities 161 Keynote Contribution: Wim Elfrink, EVP at Cisco 162 Relevance for IoT 165 Monaco Case Study 166 The City Platform 167 Crowd Management at and Around the Monaco Train Station 169 Fleet Management by Geolocalization 169 Monaco 3.0 Mobility App 169 Lessons Learned and Outlook 170 Part I Conclusions and Outlook 171 Part II. Ignite | IoT Methodology 7. Ignite | IoT Strategy Execution. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 179 IoT Strategy 182 IoT Opportunity Identification 183 IoT Opportunity Categories 184 IoT Idea Generation Process 184 Idea Refinement 186 Opportunity Qualification 189 IoT Opportunity Management 189 Business Model Development 189 Impact and Risk Analysis 198 Opportunity Selection 200 Project Initiation 202 IoT Center of Excellence 204 IoT Platform 205 Summary and Conclusions 205 8. Ignite | IoT Solution Delivery. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 207 IoT Solution Lifecycle: Plan/Build/Run 211 Assumptions 211 Approach Taken 212 IoT Project Initiation 213 Initial Solution Design 214 Make-or-Buy Decision 236 IoT Project Structure 237 Conway’s Law 237 IoT Project Workstreams 238 IoT and Agile 251 An Open Letter to the Ignite Team 251 Limitations of Agile in the IoT 252 An Agile Approach to IoT 254 Building Blocks 255 IoT Project Dimensions 256 viii TABLE OF CONTENTS IoT Architecture Blueprints 259 IoT Technology Profiles 268 Part III. Detailed Case Study 9. Background Information. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 405 Industrial Internet Consortium Testbeds 405 Track & Trace Testbed 407 Phased Approach 408 Testbed Sponsors 409 Phase 1 410 Phase 2 411 First Milestone: Bosch ConnectedWorld 2015 412 Industrial Power Tools: End-User Perspective (Airbus) 412 Industrial Power Tools: Vendor Perspective (Bosch Rexroth) 415 10. Developing Track & Trace with the Ignite | IoT Methodology. . . . . . . . . 419 IIC RA: Business Viewpoint 420 Problem Statement 421 Stakeholder Analysis 421 Site Survey 422 Project Dimensions 423 IIC RA: Usage Viewpoint 425 Use Cases 425 Solution Sketch 425 IIC RA: Functional Viewpoint 427 Getting Started: UI Mock-Ups 427 Domain Model 429 Asset Integration Architecture 430 Mapping the Domain Model to the AIA 431 SOA 432 IIC RA: Implementation Viewpoint 434 Software Architecture 434 Technical Infrastructure View 435 Results 437 11. Conclusions and Outlook. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 443 A. References. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 445 Index. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 449 TABLE OF CONTENTS ix 9789352132515

]]> Place hold on Enterprise IoT

Make : raspberry pi and avr projects

]]> Kolkata Shroff publishers & distributors 2015 .
xvii, 234 9789351109143

]]> Place hold on Make

Python for everybody : exploring data using python 3

]]> By Serverance, Charles R. .
Kolkata Shroff publishers & distributors 2009 .
xii, 235 9789352136278

]]> Place hold on Python for everybody

Practical time series analysis : prediction with statistics and machine learning

]]> By Nielsen, Aileen .
Kolkata Shroff publishers & distributors 2020 .
xvi, 480 , Table of Contents Preface. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ix 1. Time Series: An Overview and a Quick History. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 The History of Time Series in Diverse Applications 2 Medicine as a Time Series Problem 2 Forecasting Weather 6 Forecasting Economic Growth 7 Astronomy 9 Time Series Analysis Takes Off 10 The Origins of Statistical Time Series Analysis 12 The Origins of Machine Learning Time Series Analysis 13 More Resources 13 2. Finding and Wrangling Time Series Data. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17 Where to Find Time Series Data 18 Prepared Data Sets 18 Found Time Series 25 Retrofitting a Time Series Data Collection from a Collection of Tables 26 A Worked Example: Assembling a Time Series Data Collection 27 Constructing a Found Time Series 33 Timestamping Troubles 35 Whose Timestamp? 35 Guesstimating Timestamps to Make Sense of Data 36 What’s a Meaningful Time Scale? 39 Cleaning Your Data 40 Handling Missing Data 40 Upsampling and Downsampling 52 Smoothing Data 55 iiiSeasonal Data 60 Time Zones 63 Preventing Lookahead 67 More Resources 69 3. Exploratory Data Analysis for Time Series. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 73 Familiar Methods 73 Plotting 74 Histograms 77 Scatter Plots 78 Time Series–Specific Exploratory Methods 81 Understanding Stationarity 82 Applying Window Functions 86 Understanding and Identifying Self-Correlation 91 Spurious Correlations 102 Some Useful Visualizations 104 1D Visualizations 104 2D Visualizations 105 3D Visualizations 113 More Resources 117 4. Simulating Time Series Data. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 119 What’s Special About Simulating Time Series? 120 Simulation Versus Forecasting 121 Simulations in Code 121 Doing the Work Yourself 122 Building a Simulation Universe That Runs Itself 128 A Physics Simulation 134 Final Notes on Simulations 140 Statistical Simulations 141 Deep Learning Simulations 141 More Resources 142 5. Storing Temporal Data. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 143 Defining Requirements 145 Live Data Versus Stored Data 146 Database Solutions 148 SQL Versus NoSQL 149 Popular Time Series Database and File Solutions 152 File Solutions 157 NumPy 158 Pandas 158 iv | Table of ContentsStandard R Equivalents 158 Xarray 159 More Resources 160 6. Statistical Models for Time Series. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 163 Why Not Use a Linear Regression? 163 Statistical Methods Developed for Time Series 166 Autoregressive Models 166 Moving Average Models 181 Autoregressive Integrated Moving Average Models 186 Vector Autoregression 196 Variations on Statistical Models 201 Advantages and Disadvantages of Statistical Methods for Time Series 203 More Resources 204 7. State Space Models for Time Series. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 207 State Space Models: Pluses and Minuses 209 The Kalman Filter 210 Overview 210 Code for the Kalman Filter 212 Hidden Markov Models 218 How the Model Works 218 How We Fit the Model 220 Fitting an HMM in Code 224 Bayesian Structural Time Series 229 Code for bsts 230 More Resources 235 8. Generating and Selecting Features for a Time Series. . . . . . . . . . . . . . . . . . . . . . . . . . . . 239 Introductory Example 240 General Considerations When Computing Features 241 The Nature of the Time Series 242 Domain Knowledge 242 External Considerations 243 A Catalog of Places to Find Features for Inspiration 243 Open Source Time Series Feature Generation Libraries 244 Domain-Specific Feature Examples 249 How to Select Features Once You Have Generated Them 252 Concluding Thoughts 255 More Resources 256 Table of Contents | v9. Machine Learning for Time Series. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 259 Time Series Classification 260 Selecting and Generating Features 260 Decision Tree Methods 264 Clustering 272 Generating Features from the Data 273 Temporally Aware Distance Metrics 280 Clustering Code 285 More Resources 287 10. Deep Learning for Time Series. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 289 Deep Learning Concepts 292 Programming a Neural Network 294 Data, Symbols, Operations, Layers, and Graphs 294 Building a Training Pipeline 298 Inspecting Our Data Set 299 Steps of a Training Pipeline 302 Feed Forward Networks 318 A Simple Example 318 Using an Attention Mechanism to Make Feed Forward Networks More Time-Aware 321 CNNs 324 A Simple Convolutional Model 325 Alternative Convolutional Models 327 RNNs 330 Continuing Our Electric Example 332 The Autoencoder Innovation 334 Combination Architectures 335 Summing Up 340 More Resources 341 11. Measuring Error. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 343 The Basics: How to Test Forecasts 344 Model-Specific Considerations for Backtesting 347 When Is Your Forecast Good Enough? 348 Estimating Uncertainty in Your Model with a Simulation 350 Predicting Multiple Steps Ahead 353 Fit Directly to the Horizon of Interest 353 Recursive Approach to Distant Temporal Horizons 354 Multitask Learning Applied to Time Series 354 Model Validation Gotchas 355 More Resources 355 vi | Table of Contents12. Performance Considerations in Fitting and Serving Time Series Models. . . . . . . . . . . . 357 Working with Tools Built for More General Use Cases 358 Models Built for Cross-Sectional Data Don’t “Share” Data Across Samples 358 Models That Don’t Precompute Create Unnecessary Lag Between Measuring Data and Making a Forecast 360 Data Storage Formats: Pluses and Minuses 361 Store Your Data in a Binary Format 361 Preprocess Your Data in a Way That Allows You to “Slide” Over It 362 Modifying Your Analysis to Suit Performance Considerations 362 Using All Your Data Is Not Necessarily Better 363 Complicated Models Don’t Always Do Better Enough 363 A Brief Mention of Alternative High-Performance Tools 364 More Resources 365 13. Healthcare Applications. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 367 Predicting the Flu 367 A Case Study of Flu in One Metropolitan Area 367 What Is State of the Art in Flu Forecasting? 383 Predicting Blood Glucose Levels 384 Data Cleaning and Exploration 385 Generating Features 390 Fitting a Model 396 More Resources 401 14. Financial Applications. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 403 Obtaining and Exploring Financial Data 404 Preprocessing Financial Data for Deep Learning 410 Adding Quantities of Interest to Our Raw Values 410 Scaling Quantities of Interest Without a Lookahead 411 Formatting Our Data for a Neural Network 413 Building and Training an RNN 416 More Resources 423 15. Time Series for Government. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 425 Obtaining Governmental Data 426 Exploring Big Time Series Data 428 Upsample and Aggregate the Data as We Iterate Through It 431 Sort the Data 432 Online Statistical Analysis of Time Series Data 436 Remaining Questions 446 Further Improvements 446 More Resources 447 Table of Contents | vii16. Time Series Packages. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 449 Forecasting at Scale 449 Google’s Industrial In-house Forecasting 450 Facebook’s Open Source Prophet Package 452 Anomaly Detection 457 Twitter’s Open Source AnomalyDetection Package 457 Other Time Series Packages 460 More Resources 461 17. Forecasts About Forecasting. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 463 Forecasting as a Service 463 Deep Learning Enhances Probabilistic Possibilities 464 Increasing Importance of Machine Learning Rather Than Statistics 465 Increasing Combination of Statistical and Machine Learning Methodologies 466 More Forecasts for Everyday Life 466 Index. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 469 9789352139255

]]> Place hold on Practical time series analysis

Python : for beginners

]]> By Borate, Rahul E. .
Kolkata Shroff publishers & distributors 2019 .
xiii, 137 , Table of Content CHAPTER 1: INTRODUCTION TO PYTHON .......................................1 1.1 Getting Started: Introduction to Python- an interpreted high level language, interactive mode and script mode .................................................. 2 1.2 Variables and Types-mutable and Immutable variable and Keywords ...... 6 1.3 Operators and Operands in Python .................................................................. 7 1.4 Operator precedence, Expressions and Statements (Assignment statement) .....................................................................................11 1.5 Taking input (using raw_input() and input()) and displaying output - print statement ................................................................ 12 1.6 Comments in Python ......................................................................................... 13 Exercise ................................................................................................................ 14 CHAPTER 2: CONDITIONAL AND LOOPING CONSTRUCT ..............15 2.1 if - else statement and nested if – else while, for, use of range function in for, Nested loops ............................................................................ 16 2.2 break, continue, pass statement ....................................................................... 24 2.3 Use of Compound Expression in conditional construct .............................. 26 2.4 Built-In Function, invoking built in functions ............................................... 27 2.5 Functions from Math, Random, Time & Date Module ................................ 29 2.6 Module (Importing entire module or selected objects using from statement) .......................................................................... 33 2.7 Composition ....................................................................................................... 33 2.8 User Defined Function: Defining, invoking functions, passing parameters (default parameter values, keyword arguments) ......................................... 34 2.9 Scope of Variables, Void Functions and Functions Returning Values Scope of Variables ................................................................................. 38 Exercise ................................................................................................................ 43 xii Python Programming for Beginners CHAPTER 3: STRINGS .......................................................45 3.1 Creating, Initializing and Accessing the Elements ....................................... 46 3.2 String Operators ................................................................................................. 48 3.3 String built in functions & methods: ............................................................... 51 Exercise ................................................................................................................ 67 CHAPTER 4: STRUCTURED PROGRAMMING: ITERATION CONTROL FLOW ...........................................................69 4.1 Concept of Mutable Lists, Creating, Initializing and Accessing the Elements of List ........................................................................ 70 4.2 List Operations (Concatenation, Repetition, Membership, List Slices), List Comprehensions .................................................................... 72 4.3 List Functions & Methods: append, extend, sort, remove, reverse, pop ......................................................................................... 74 4.4 Immutable concept, creating, initializing and accessing the elements in a tuple ..................................................................... 78 4.5 Tuple Functions: cmp(), len(), max(), min(), tuple() ..................................... 79 4.6 Concept of Sets, Creating, Initializing and Accessing the Elements of Sets ........................................................................ 82 4.7 Sets Operation (Membership, Union, Intersection, Difference, and Symmetric Difference) ............................................................................... 83 4.8 Concept of Key-Value Pair, Creating, Initializing and Accessing the Elements in a Dictionary ......................................................... 85 4.9 Traversing, Appending, Updating and Deleting Elements ......................... 87 4.10 Dictionary Functions & Methods: cmp, len, clear(), has_key(), items(), keys(), update(), values() ................................................. 88 Exercise ................................................................................................................ 92 CHAPTER 5: MODULES .....................................................................93 5.1 Concept of Module: Executing Modules as Scripts, the Module Search Path, “Compiled” Python Files, Standard Modules: What is Module? ............................................................. 94 Table of Content xiii 5.2 The dir() Function ................................................................................................ 97 5.3 Package ................................................................................................................. 98 Exercise ...................................................................................................................... 101 CHAPTER 6: I/O AND FILE HANDLING .........................................103 6.1 Output Formatting: .......................................................................................... 104 6.2 Filenames and Paths:........................................................................................ 106 6.3 Reading and Writing Files:.............................................................................. 109 Exercise:.......................................................................................................................113 CHAPTER 7: ERRORS AND EXCEPTIONS .......................................115 7.1 Syntax Errors, Exceptions:................................................................................116 7.2 Handling Exceptions:........................................................................................118 7.3 Python Exception(Except with No Exception) Example: .......................... 120 7.4 Raise an Exception:........................................................................................... 123 7.5 User-defined Exceptions:................................................................................. 124 7.6 Clean-Up Actions (Try ... Finally):.................................................................. 125 Exercise:...................................................................................................................... 128 CHAPTER 8: INTRODUCTION TO OBJECT ORIENTED CONCEPTS IN PYTHON ..................................................................129 8.1 Object Oriented concepts ................................................................................ 130 8.3 Classes, Class Objects, Instance Objects, Method Objects, Class and Instance Variables: Class and Instance Variables in Python ...................... 131 8.4 Inheritance ........................................................................................................ 134 Exercise ...................................................................................................................... 137 9789352138753

]]> Place hold on Python

Python : for beginners

]]> Place hold on Python

Head first python

]]> By Barry, Paul .
Kolkata Shroff publishers & distributors 2017 .
xxxviii, 584 , Includes appendix & index Table of Contents (Summary) 1 The Basics: Getting Started Quickly 1 2 List Data: Working with Ordered Data 47 3 Structured Data: Working with Structured Data 95 4 Code Reuse: Functions and Modules 145 5 Building a Webapp: Getting Real 195 6 Storing and Manipulating Data: Where to Put Your Data 243 7 Using a Database: Putting Python’s DB-API to Use 281 8 A Little Bit of Class: Abstracting Behavior and State 309 9 The Context Management Protocol: Hooking into Python’s with Statement 335 10 Function Decorators: Wrapping Functions 363 11 Exception Handling: What to Do When Things Go Wrong 413 11¾ A Little Bit of Threading: Dealing with Waiting 461 12 Advanced Iteration: Looping like Crazy 477 A Installing: Installing Python 521 B Pythonanywhere: Deploying Your Webapp 529 C Top Ten Things We Didn’t Cover: There’s Always More to Learn 539 D Top Ten Projects Not Covered: Even More Tools, Libraries, and Modules 551 E Getting Involved: The Python Community 563 Table of Contents (the real thing) Your brain on Python. Here you are trying to learn something, while here your brain is, doing you a favor by making sure the learning doesn’t stick. Your brain’s thinking, “Better leave room for more important things, like which wild animals to avoid and whether naked snowboarding is a bad idea.” So how do you trick your brain into thinking that your life depends on knowing how to program in Python? Intro Who is this book for? xxviii We know what you’re thinking xxix We know what your brain is thinking xxix Metacognition: thinking about thinking xxxi Here’s what WE did xxxii Read me xxxiv Acknowledgments xxxvii table of contents x the basics Getting Started Quickly Get going with Python programming as quickly as possible. In this chapter, we introduce the basics of programming in Python, and we do this in typical Head First style: by jumping right in. After just a few pages, you’ll have run your first sample program. By the end of the chapter, you’ll not only be able to run the sample program, but you’ll understand its code too (and more besides). Along the way, you’ll learn about a few of the things that make Python the programming language it is. Understanding IDLE’s Windows 4 Executing Code, One Statement at a Time 8 Functions + Modules = The Standard Library 9 Data Structures Come Built-in 13 Invoking Methods Obtains Results 14 Deciding When to Run Blocks of Code 15 What “else” Can You Have with “if ”? 17 Suites Can Contain Embedded Suites 18 Returning to the Python Shell 22 Experimenting at the Shell 23 Iterating Over a Sequence of Objects 24 Iterating a Specific Number of Times 25 Applying the Outcome of Task #1 to Our Code 26 Arranging to Pause Execution 28 Generating Random Integers with Python 30 Coding a Serious Business Application 38 Is Indentation Driving You Crazy? 40 Asking the Interpreter for Help on a Function 41 Experimenting with Ranges 42 Chapter 1’s Code 46 1 table of contents xi list data Working with Data All programs process data, and Python programs are no exception. In fact, take a look around: data is everywhere. A lot of, if not most, programming is all about data: acquiring data, processing data, understanding data. To work with data effectively, you need somewhere to put your data when processing it. Python shines in this regard, thanks (in no small part) to its inclusion of a handful of widely applicable data structures: lists, dictionaries, tuples, and sets. In this chapter, we’ll preview all four, before spending the majority of this chapter digging deeper into lists (and we’ll deep-dive into the other three in the next chapter). We’re covering these data structures early, as most of what you’ll likely do with Python will revolve around working with data. 0 D -12 1 o -11 2 n -10 3 ' -9 4 t -8 5 -7 6 p -6 7 a -5 8 n -4 9 i -3 10 c -2 11 ! -1 Numbers, Strings...and Objects 48 Meet the Four Built-in Data Structures 50 An Unordered Data Structure: Dictionary 52 A Data Structure That Avoids Duplicates: Set 53 Creating Lists Literally 55 Use Your Editor When Working on More Than a Few Lines of Code 57 “Growing” a List at Runtime 58 Checking for Membership with “in” 59 Removing Objects from a List 62 Extending a List with Objects 64 Inserting an Object into a List 65 How to Copy a Data Structure 73 Lists Extend the Square Bracket Notation 75 Lists Understand Start, Stop, and Step 76 Starting and Stopping with Lists 78 Putting Slices to Work on Lists 80 Python’s “for” Loop Understands Lists 86 Marvin’s Slices in Detail 88 When Not to Use Lists 91 Chapter 2’s Code, 1 of 2 92 2 table of contents xii Name: Ford Prefect Gender: Male Occupation: Researcher Home Planet: Betelgeuse Seven structured data Working with Structured Data Python’s list data structure is great, but it isn’t a data panacea. When you have truly structured data (and using a list to store it may not be the best choice), Python comes to your rescue with its built-in dictionary. Out of the box, the dictionary lets you store and manipulate any collection of key/value pairs. We look long and hard at Python’s dictionary in this chapter, and—along the way—meet set and tuple, too. Together with the list (which we met in the previous chapter), the dictionary, set, and tuple data structures provide a set of built-in data tools that help to make Python and data a powerful combination. A Dictionary Stores Key/Value Pairs 96 How to Spot a Dictionary in Code 98 Insertion Order Is NOT Maintained 99 Value Lookup with Square Brackets 100 Working with Dictionaries at Runtime 101 Updating a Frequency Counter 105 Iterating Over a Dictionary 107 Iterating Over Keys and Values 108 Iterating Over a Dictionary with “items” 110 Just How Dynamic Are Dictionaries? 114 Avoiding KeyErrors at Runtime 116 Checking for Membership with “in” 117 Ensuring Initialization Before Use 118 Substituting “not in” for “in” 119 Putting the “setdefault” Method to Work 120 Creating Sets Efficiently 124 Taking Advantage of Set Methods 125 Making the Case for Tuples 132 Combining the Built-in Data Structures 135 Accessing a Complex Data Structure’s Data 141 Chapter 3’s Code, 1 of 2 143 3 table of contents xiii module code reuse Functions and Modules Reusing code is key to building a maintainable system. And when it comes to reusing code in Python, it all starts and ends with the humble function. Take some lines of code, give them a name, and you’ve got a function (which can be reused). Take a collection of functions and package them as a file, and you’ve got a module (which can also be reused). It’s true what they say: it’s good to share, and by the end of this chapter, you’ll be well on your way to sharing and reusing your code, thanks to an understanding of how Python’s functions and modules work. Reusing Code with Functions 146 Introducing Functions 147 Invoking Your Function 150 Functions Can Accept Arguments 154 Returning One Value 158 Returning More Than One Value 159 Recalling the Built-in Data Structures 161 Making a Generically Useful Function 165 Creating Another Function, 1 of 3 166 Specifying Default Values for Arguments 170 Positional Versus Keyword Assignment 171 Updating What We Know About Functions 172 Running Python from the Command Line 175 Creating the Required Setup Files 179 Creating the Distribution File 180 Installing Packages with “pip” 182 Demonstrating Call-by-Value Semantics 185 Demonstrating Call-by-Reference Semantics 186 Install the Testing Developer Tools 190 How PEP 8–Compliant Is Our Code? 191 Understanding the Failure Messages 192 Chapter 4’s Programs 194 4 table of contents xiv building a webapp Getting Real At this stage, you know enough Python to be dangerous. With this book’s first four chapters behind you, you’re now in a position to productively use Python within any number of application areas (even though there’s still lots of Python to learn). Rather than explore the long list of what these application areas are, in this and subsequent chapters, we’re going to structure our learning around the development of a web-hosted application, which is an area where Python is especially strong. Along the way, you’ll learn a bit more about Python. Python: What You Already Know 196 What Do We Want Our Webapp to Do? 200 Let’s Install Flask 202 How Does Flask Work? 203 Running Your Flask Webapp for the First Time 204 Creating a Flask Webapp Object 206 Decorating a Function with a URL 207 Running Your Webapp’s Behavior(s) 208 Exposing Functionality to the Web 209 Building the HTML Form 213 Templates Relate to Web Pages 216 Rendering Templates from Flask 217 Displaying the Webapp’s HTML Form 218 Preparing to Run the Template Code 219 Understanding HTTP Status Codes 222 Handling Posted Data 223 Refining the Edit/Stop/Start/Test Cycle 224 Accessing HTML Form Data with Flask 226 Using Request Data in Your Webapp 227 Producing the Results As HTML 229 Preparing Your Webapp for the Cloud 238 Chapter 5’s Code 241 5 table of contents xv Form Data Remote_addr User_agent Results ImmutableMultiDict([(‘phrase’, 127.0.0.1 Mozilla/5.0 (Macintosh; {‘e’, ‘i’} ‘hitch-hiker’), (‘letters’, ‘aeiou’)]) Intel Mac OS X 10_11_2) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/47.0.2526 .106 Safari/537.36 storing and manipulating data Where to Put Your Data Sooner or later, you’ll need to safely store your data somewhere. And when it comes to storing data, Python has you covered. In this chapter, you’ll learn about storing and retrieving data from text files, which—as storage mechanisms go—may feel a bit simplistic, but is nevertheless used in many problem areas. As well as storing and retrieving your data from files, you’ll also learn some tricks of the trade when it comes to manipulating data. We’re saving the “serious stuff” (storing data in a database) until the next chapter, but there’s plenty to keep us busy for now when working with files. Doing Something with Your Webapp’s Data 244 Python Supports Open, Process, Close 245 Reading Data from an Existing File 246 A Better Open, Process, Close: “with” 248 View the Log Through Your Webapp 254 Examine the Raw Data with View Source 256 It’s Time to Escape (Your Data) 257 Viewing the Entire Log in Your Webapp 258 Logging Specific Web Request Attributes 261 Log a Single Line of Delimited Data 262 From Raw Data to Readable Output 265 Generate Readable Output With HTML 274 Embed Display Logic in Your Template 275 Producing Readable Output with Jinja2 276 The Current State of Our Webapp Code 278 Asking Questions of Your Data 279 Chapter 6’s Code 280 6 table of contents xvi Python’s DB-API The MySQLConnector/Python Driver MySQL Your code using a database Putting Python’s DB-API to Use Storing data in a relational database system is handy. In this chapter, you’ll learn how to write code that interacts with the popular MySQL database technology, using a generic database API called DB-API. The DB-API (which comes standard with every Python install) allows you to write code that is easily transferred from one database product to the next... assuming your database talks SQL. Although we’ll be using MySQL, there’s nothing stopping you from using your DB-API code with your favorite relational database, whatever it may be. Let’s see what’s involved in using a relational database with Python. There’s not a lot of new Python in this chapter, but using Python to talk to databases is a big deal, so it’s well worth learning. Database-Enabling Your Webapp 282 Task 1: Install the MySQL Server 283 Introducing Python’s DB-API 284 Task 2: Install a MySQL Database Driver for Python 285 Install MySQL-Connector/Python 286 Task 3: Create Our Webapp’s Database and Tables 287 Decide on a Structure for Your Log Data 288 Confirm Your Table Is Ready for Data 289 Task 4: Create Code to Work with Our Webapp’s Database and Tables 296 Storing Data Is Only Half the Battle 300 How Best to Reuse Your Database Code? 301 Consider What You’re Trying to Reuse 302 What About That Import? 303 You’ve Seen This Pattern Before 305 The Bad News Isn’t Really All That Bad 306 Chapter 7’s Code 307 7 table of contents xvii a little bit of class Abstracting Behavior and State Classes let you bundle code behavior and state together. In this chapter, you’re setting your webapp aside while you learn about creating Python classes. You’re doing this in order to get to the point where you can create a context manager with the help of a Python class. As creating and using classes is such a useful thing to know about anyway, we’re dedicating this chapter to them. We won’t cover everything about classes, but we’ll touch on all the bits you’ll need to understand in order to confidently create the context manager your webapp is waiting for. Hooking into the “with” Statement 310 An Object-Oriented Primer 311 Creating Objects from Classes 312 Objects Share Behavior but Not State 313 Doing More with CountFromBy 314 Invoking a Method: Understand the Details 316 Adding a Method to a Class 318 The Importance of “self ” 320 Coping with Scoping 321 Prefix Your Attribute Names with “self ” 322 Initialize (Attribute) Values Before Use 323 Dunder “init” Initializes Attributes 324 Initializing Attributes with Dunder “init” 325 Understanding CountFromBy’s Representation 328 Defining CountFromBy’s Representation 329 Providing Sensible Defaults for CountFromBy 330 Classes: What We Know 332 Chapter 8’s Code 333 8 table of contents xviii § $ mysql -u vsearch -p vsearchlogDB Enter password: Welcome to MySQL monitor... mysql> select * from log; +----+---------------------+--------------------------+---------+-----------+----------------+----------------------+ | id | ts | phrase | letters | ip | browser_string | results | +----+---------------------+--------------------------+---------+-----------+----------------+----------------------+ | 1 | 2016-03-09 13:40:46 | life, the uni ... ything | aeiou | 127.0.0.1 | firefox | {'u', 'e', 'i', 'a'} | | 2 | 2016-03-09 13:42:07 | hitch-hiker | aeiou | 127.0.0.1 | safari | {'i', 'e'} | | 3 | 2016-03-09 13:42:15 | galaxy | xyz | 127.0.0.1 | chrome | {'y', 'x'} | | 4 | 2016-03-09 13:43:07 | hitch-hiker | xyz | 127.0.0.1 | firefox | set() | +----+---------------------+--------------------------+---------+-----------+----------------+----------------------+ 4 rows in set (0.0 sec) mysql> quit Bye File Edit Window Help Checking our log DB the context management protocol Hooking into Python’s with Statements It’s time to take what you’ve just learned and put it to work. Chapter 7 discussed using a relational database with Python, while Chapter 8 provided an introduction to using classes in your Python code. In this chapter, both of these techniques are combined to produce a context manager that lets us extend the with statement to work with relational database systems. In this chapter, you’ll hook into the with statement by creating a new class, which conforms to Python’s context management protocol. What’s the Best Way to Share Our Webapp’s Database Code? 336 Managing Context with Methods 338 You’ve Already Seen a Context Manager in Action 339 Create a New Context Manager Class 340 Initialize the Class with the Database Config 341 Perform Setup with Dunder “enter” 343 Perform Teardown with Dunder “exit” 345 Reconsidering Your Webapp Code, 1 of 2 348 Recalling the “log_request” Function 350 Amending the “log_request” Function 351 Recalling the “view_the_log” Function 352 It’s Not Just the Code That Changes 353 Amending the “view_the_log” Function 354 Answering the Data Questions 359 Chapter 9’s Code, 1 of 2 360 9 table of contents xix function decorators Wrapping Functions When it comes to augmenting your code, Chapter 9’s context management protocol is not the only game in town. Python also lets you use function decorators, a technique whereby you can add code to an existing function without having to change any of the existing function’s code. If you think this sounds like some sort of black art, don’t despair: it’s nothing of the sort. However, as coding techniques go, creating a function decorator is often considered to be on the harder side by many Python programmers, and thus is not used as often as it should be. In this chapter, our plan is to show you that, despite being an advanced technique, creating and using your own decorators is not that hard. Your Web Server (Not Your Computer) Runs Your Code 366 Flask’s Session Technology Adds State 368 Dictionary Lookup Retrieves State 369 Managing Logins with Sessions 374 Let’s Do Logout and Status Checking 377 Pass a Function to a Function 386 Invoking a Passed Function 387 Accepting a List of Arguments 390 Processing a List of Arguments 391 Accepting a Dictionary of Arguments 392 Processing a Dictionary of Arguments 393 Accepting Any Number and Type of Function Arguments 394 Creating a Function Decorator 397 The Final Step: Handling Arguments 401 Putting Your Decorator to Work 404 Back to Restricting Access to /viewlog 408 Chapter 10’s Code, 1 of 2 410 10 table of contents xx ... Exception +-- StopIteration +-- StopAsyncIteration +-- ArithmeticError | +-- FloatingPointError | +-- OverflowError | +-- ZeroDivisionError +-- AssertionError +-- AttributeError +-- BufferError +-- EOFError ... exception handling What to Do When Things Go Wrong Things go wrong, all the time—no matter how good your code is. You’ve successfully executed all of the examples in this book, and you’re likely confident all of the code presented thus far works. But does this mean the code is robust? Probably not. Writing code based on the assumption that nothing bad ever happens is (at best) naive. At worst, it’s dangerous, as unforeseen things do (and will) happen. It’s much better if you’re wary while coding, as opposed to trusting. Care is needed to ensure your code does what you want it to, as well as reacts properly when things go south. Databases Aren’t Always Available 418 Web Attacks Are a Real Pain 419 Input-Output Is (Sometimes) Slow 420 Your Function Calls Can Fail 421 Always Try to Execute Error-Prone Code 423 try Once, but except Many Times 426 The Catch-All Exception Handler 428 Learning About Exceptions from “sys” 430 The Catch-All Exception Handler, Revisited 431 Getting Back to Our Webapp Code 433 Silently Handling Exceptions 434 Handling Other Database Errors 440 Avoid Tightly Coupled Code 442 The DBcm Module, Revisited 443 Creating Custom Exceptions 444 What Else Can Go Wrong with “DBcm”? 448 Handling SQLError Is Different 451 Raising an SQLError 453 A Quick Recap: Adding Robustness 455 How to Deal with Wait? It Depends... 456 Chapter 11’s Code, 1 of 3 457 11 table of contents xxi Wait! a little bit of threading Dealing with Waiting Your code can sometimes take a long time to execute. Depending on who notices, this may or may not be an issue. If some code takes 30 seconds to do its thing “behind the scenes,” the wait may not be an issue. However, if your user is waiting for your application to respond, and it takes 30 seconds, everyone notices. What you should do to fix this problem depends on what you’re trying to do (and who’s doing the waiting). In this short chapter, we’ll briefly discuss some options, then look at one solution to the issue at hand: what happens if something takes too long? Waiting: What to Do? 462 How Are You Querying Your Database? 463 Database INSERTs and SELECTs Are Different 464 Doing More Than One Thing at Once 465 Don’t Get Bummed Out: Use Threads 466 First Things First: Don’t Panic 470 Don’t Get Bummed Out: Flask Can Help 471 Is Your Webapp Robust Now? 474 Chapter 11¾’s Code, 1 of 2 475 113 /4 table of contents xxii advanced iteration Looping Like Crazy It’s often amazing how much time our programs spend in loops. This isn’t a surprise, as most programs exist to perform something quickly a whole heap of times. When it comes to optimizing loops, there are two approaches: (1) improve the loop syntax (to make it easier to specify a loop), and (2) improve how loops execute (to make them go faster). Early in the lifetime of Python 2 (that is, a long, long time ago), the language designers added a single language feature that implements both approaches, and it goes by a rather strange name: comprehension. Reading CSV Data As Lists 479 Reading CSV Data As Dictionaries 480 Stripping, Then Splitting, Your Raw Data 482 Be Careful When Chaining Method Calls 483 Transforming Data into the Format You Need 484 Transforming into a Dictionary Of Lists 485 Spotting the Pattern with Lists 490 Converting Patterns into Comprehensions 491 Take a Closer Look at the Comprehension 492 Specifying a Dictionary Comprehension 494 Extend Comprehensions with Filters 495 Deal with Complexity the Python Way 499 The Set Comprehension in Action 505 What About “Tuple Comprehensions”? 507 Parentheses Around Code == Generator 508 Using a Listcomp to Process URLs 509 Using a Generator to Process URLs 510 Define What Your Function Needs to Do 512 Yield to the Power of Generator Functions 513 Tracing Your Generator Function, 1 of 2 514 One Final Question 518 Chapter 12’s Code 519 It’s Time to Go… 520 12 table of contents xxiii installation Installing Python pythonanywhere Deploying Your Webapp First things first: let’s get Python installed on your computer. Whether you’re running on Windows, Mac OS X, or Linux, Python’s got you covered. How you install it on each of these platforms is specific to how things work on each of these operating systems (we know...a shocker, eh?), and the Python community works hard to provide installers that target all the popular systems. In this short appendix, you’ll be guided through installing Python on your computer. At the end of Chapter 5, we claimed that deploying your webapp to the cloud was only 10 minutes away. It’s now time to make good on that promise. In this appendix, we are going to take you through the process of deploying your webapp on PythonAnywhere, going from zero to deployed in about 10 minutes. PythonAnywhere is a favorite among the Python programming community, and it’s not hard to see why: it works exactly as you’d expect it to, has great support for Python (and Flask), and—best of all—you can get started hosting your webapp at no cost. Install Python 3 on Windows 522 Check Python 3 on Windows 523 Add to Python 3 on Windows 524 Install Python 3 on Mac OS X (macOS) 525 Check and Configure Python 3 on Mac OS X 526 Install Python 3 on Linux 527 Step 0: A Little Prep 530 Step 1: Sign Up for PythonAnywhere 531 Step 2: Upload Your Files to the Cloud 532 Step 3: Extract and Install Your Code 533 Step 4: Create a Starter Webapp, 1 of 2 534 Step 5: Configure Your Webapp 536 Step 6: Take Your Cloud-Based Webapp for a Spin! 537 a b table of contents xxiv top ten things we didn’t cover There’s Always More to Learn It was never our intention to try to cover everything. This book’s goal was always to show you enough Python to get you up to speed as quickly as possible. There’s a lot more we could’ve covered, but didn’t. In this appendix, we discuss the top 10 things that—given another 600 pages or so—we would’ve eventually gotten around to. Not all of the 10 things will interest you, but quickly flip through them just in case we’ve hit on your sweet spot, or provided an answer to that nagging question. All the programming technologies in this appendix come baked in to Python and its interpreter. 1. What About Python 2? 540 2. Virtual Programming Environments 541 3. More on Object Orientation 542 4. Formats for Strings and the Like 543 5. Getting Things Sorted 544 6. More from the Standard Library 545 7. Running Your Code Concurrently 546 8. GUIs with Tkinter (and Fun with Turtles) 547 9. It’s Not Over ’Til It’s Tested 548 10. Debug, Debug, Debug 549 c table of contents xxv top ten projects not covered Even More Tools, Libraries, and Modules We know what you’re thinking as you read this appendix’s title. Why on Earth didn’t they make the title of the last appendix: The Top Twenty Things We Didn’t Cover? Why another 10? In the last appendix, we limited our discussion to stuff that comes baked in to Python (part of the language’s “batteries included”). In this appendix, we cast the net much further afield, discussing a whole host of technologies that are available to you because Python exists. There’s lots of good stuff here and—just like with the last appendix—a quick perusal won’t hurt you one single bit. 1. Alternatives to >>> 552 2. Alternatives to IDLE 553 3. Jupyter Notebook: The Web-Based IDE 554 4. Doing Data Science 555 5. Web Development Technologies 556 6. Working with Web Data 557 7. More Data Sources 558 8. Programming Tools 559 9. Kivy: Our Pick for “Coolest Project Ever” 560 10. Alternative Implementations 561 d table of contents xxvi getting involved The Python Community Python is much more than a great programming language. It’s a great community, too. The Python Community is welcoming, diverse, open, friendly, sharing, and giving. We’re just amazed that no one, to date, has thought to put that on a greeting card! Seriously, though, there’s more to programming in Python than the language. An entire ecosystem has grown up around Python, in the form of excellent books, blogs, websites, conferences, meetups, user groups, and personalities. In this appendix, we take a survey of the Python community and see what it has to offer. Don’t just sit around programming on your own: get involved! BDFL: Benevolent Dictator for Life 564 A Tolerant Community: Respect for Diversity 565 Python Podcasts 566 The Zen of Python 567 Which Book Should I Read Next? 568 Our Favorite Python Books 569 9789352134823

]]> Place hold on Head first python

R in a nutshell

]]> By Adler, Joseph .
Kolkata Shroff publishers & distributors 2012 .
xix, 699 , Includes index 9789350239209

]]> Place hold on R in a nutshell

Data science for business

]]> By Provost, Foster .
New Delhi Shroff publishers & distributors pvt. ltd. 2013 .
xviii, 384 , Includes index, glossary & bibliography 9789351102670

]]> Place hold on Data science for business

Deep Learning : A practitioner's approach

]]> By Patterson, Josh .
New Delhi Shroff publishers & distributors 2021 .
xxi, 507 , Includes index & appendix 9789352136049

]]> Place hold on Deep Learning

Practical machine learning for computer vision : end-to end machine learning for images

]]> By Lakshmanan, Valliappa .
Kolkata Shroff publishers & distributors 2021 .
xvi, 463 , Includes index 9789391043834

]]> Place hold on Practical machine learning for computer vision

Analytical skills for AI and data science : building skills for an AI- Driven enterprise

]]> By Vaughan, Daniel.
New Delhi Shroff publishers & Distributors 2020 .
xii, 228 , Includes index 9789352139835

]]> Place hold on Analytical skills for AI and data science

Hands-on machine learning with scikit-learn, keras, and tensorflow : concepts, tools, and techniques to build intelligent systems

]]> By Geron, Aurelien .
New Delhi Shroff publishers & distributors 2022 .
xxv, 834 , Part I. The Fundamentals of Machine Learning 1. The Machine Learning Landscape. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 What Is Machine Learning? 2 Why Use Machine Learning? 2 Examples of Applications 5 Types of Machine Learning Systems 7 Supervised/Unsupervised Learning 7 Batch and Online Learning 14 Instance-Based Versus Model-Based Learning 17 Main Challenges of Machine Learning 23 Insufficient Quantity of Training Data 23 Nonrepresentative Training Data 25 Poor-Quality Data 26 Irrelevant Features 27 Overfitting the Training Data 27 Underfitting the Training Data 29 Stepping Back 30 Testing and Validating 30 Hyperparameter Tuning and Model Selection 31 Data Mismatch 32 Exercises 33 2. End-to-End Machine Learning Project. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35 Working with Real Data 35 iii Look at the Big Picture 37 Frame the Problem 37 Select a Performance Measure 39 Check the Assumptions 42 Get the Data 42 Create the Workspace 42 Download the Data 46 Take a Quick Look at the Data Structure 47 Create a Test Set 51 Discover and Visualize the Data to Gain Insights 56 Visualizing Geographical Data 56 Looking for Correlations 58 Experimenting with Attribute Combinations 61 Prepare the Data for Machine Learning Algorithms 62 Data Cleaning 63 Handling Text and Categorical Attributes 65 Custom Transformers 68 Feature Scaling 69 Transformation Pipelines 70 Select and Train a Model 72 Training and Evaluating on the Training Set 72 Better Evaluation Using Cross-Validation 73 Fine-Tune Your Model 75 Grid Search 76 Randomized Search 78 Ensemble Methods 78 Analyze the Best Models and Their Errors 78 Evaluate Your System on the Test Set 79 Launch, Monitor, and Maintain Your System 80 Try It Out! 83 Exercises 84 3. Classification. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 85 MNIST 85 Training a Binary Classifier 88 Performance Measures 88 Measuring Accuracy Using Cross-Validation 89 Confusion Matrix 90 Precision and Recall 92 Precision/Recall Trade-off 93 The ROC Curve 97 Multiclass Classification 100 iv | Table of Contents Error Analysis 102 Multilabel Classification 106 Multioutput Classification 107 Exercises 108 4. Training Models. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 111 Linear Regression 112 The Normal Equation 114 Computational Complexity 117 Gradient Descent 118 Batch Gradient Descent 121 Stochastic Gradient Descent 124 Mini-batch Gradient Descent 127 Polynomial Regression 128 Learning Curves 130 Regularized Linear Models 134 Ridge Regression 135 Lasso Regression 137 Elastic Net 140 Early Stopping 141 Logistic Regression 142 Estimating Probabilities 143 Training and Cost Function 144 Decision Boundaries 145 Softmax Regression 148 Exercises 151 5. Support Vector Machines. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 153 Linear SVM Classification 153 Soft Margin Classification 154 Nonlinear SVM Classification 157 Polynomial Kernel 158 Similarity Features 159 Gaussian RBF Kernel 160 Computational Complexity 162 SVM Regression 162 Under the Hood 164 Decision Function and Predictions 165 Training Objective 166 Quadratic Programming 167 The Dual Problem 168 Kernelized SVMs 169 Table of Contents | v Online SVMs 172 Exercises 174 6. Decision Trees. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 175 Training and Visualizing a Decision Tree 175 Making Predictions 176 Estimating Class Probabilities 178 The CART Training Algorithm 179 Computational Complexity 180 Gini Impurity or Entropy? 180 Regularization Hyperparameters 181 Regression 183 Instability 185 Exercises 186 7. Ensemble Learning and Random Forests. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 189 Voting Classifiers 189 Bagging and Pasting 192 Bagging and Pasting in Scikit-Learn 194 Out-of-Bag Evaluation 195 Random Patches and Random Subspaces 196 Random Forests 197 Extra-Trees 198 Feature Importance 198 Boosting 199 AdaBoost 200 Gradient Boosting 203 Stacking 208 Exercises 211 8. Dimensionality Reduction. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 213 The Curse of Dimensionality 214 Main Approaches for Dimensionality Reduction 215 Projection 215 Manifold Learning 218 PCA 219 Preserving the Variance 219 Principal Components 220 Projecting Down to d Dimensions 221 Using Scikit-Learn 222 Explained Variance Ratio 222 Choosing the Right Number of Dimensions 223 vi | Table of Contents PCA for Compression 224 Randomized PCA 225 Incremental PCA 225 Kernel PCA 226 Selecting a Kernel and Tuning Hyperparameters 227 LLE 230 Other Dimensionality Reduction Techniques 232 Exercises 233 9. Unsupervised Learning Techniques. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 235 Clustering 236 K-Means 238 Limits of K-Means 248 Using Clustering for Image Segmentation 249 Using Clustering for Preprocessing 251 Using Clustering for Semi-Supervised Learning 253 DBSCAN 255 Other Clustering Algorithms 258 Gaussian Mixtures 260 Anomaly Detection Using Gaussian Mixtures 266 Selecting the Number of Clusters 267 Bayesian Gaussian Mixture Models 270 Other Algorithms for Anomaly and Novelty Detection 274 Exercises 275 Part II. Neural Networks and Deep Learning 10. Introduction to Artificial Neural Networks with Keras. . . . . . . . . . . . . . . . . . . . . . . . . . 279 From Biological to Artificial Neurons 280 Biological Neurons 281 Logical Computations with Neurons 283 The Perceptron 284 The Multilayer Perceptron and Backpropagation 289 Regression MLPs 292 Classification MLPs 294 Implementing MLPs with Keras 295 Installing TensorFlow 2 296 Building an Image Classifier Using the Sequential API 297 Building a Regression MLP Using the Sequential API 307 Building Complex Models Using the Functional API 308 Using the Subclassing API to Build Dynamic Models 313 Table of Contents | vii Saving and Restoring a Model 314 Using Callbacks 315 Using TensorBoard for Visualization 317 Fine-Tuning Neural Network Hyperparameters 320 Number of Hidden Layers 323 Number of Neurons per Hidden Layer 325 Learning Rate, Batch Size, and Other Hyperparameters 325 Exercises 327 11. Training Deep Neural Networks. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 331 The Vanishing/Exploding Gradients Problems 332 Glorot and He Initialization 333 Nonsaturating Activation Functions 335 Batch Normalization 338 Gradient Clipping 345 Reusing Pretrained Layers 345 Transfer Learning with Keras 347 Unsupervised Pretraining 349 Pretraining on an Auxiliary Task 350 Faster Optimizers 351 Momentum Optimization 351 Nesterov Accelerated Gradient 353 AdaGrad 354 RMSProp 355 Adam and Nadam Optimization 356 Learning Rate Scheduling 359 Avoiding Overfitting Through Regularization 364 ℓ1 and ℓ2 Regularization 364 Dropout 365 Monte Carlo (MC) Dropout 368 Max-Norm Regularization 370 Summary and Practical Guidelines 371 Exercises 373 12. Custom Models and Training with TensorFlow. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 375 A Quick Tour of TensorFlow 376 Using TensorFlow like NumPy 379 Tensors and Operations 379 Tensors and NumPy 381 Type Conversions 381 Variables 382 Other Data Structures 383 viii | Table of Contents Customizing Models and Training Algorithms 384 Custom Loss Functions 384 Saving and Loading Models That Contain Custom Components 385 Custom Activation Functions, Initializers, Regularizers, and Constraints 387 Custom Metrics 388 Custom Layers 391 Custom Models 394 Losses and Metrics Based on Model Internals 397 Computing Gradients Using Autodiff 399 Custom Training Loops 402 TensorFlow Functions and Graphs 405 AutoGraph and Tracing 407 TF Function Rules 409 Exercises 410 13. Loading and Preprocessing Data with TensorFlow. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 413 The Data API 414 Chaining Transformations 415 Shuffling the Data 416 Preprocessing the Data 419 Putting Everything Together 420 Prefetching 421 Using the Dataset with tf.keras 423 The TFRecord Format 424 Compressed TFRecord Files 425 A Brief Introduction to Protocol Buffers 425 TensorFlow Protobufs 427 Loading and Parsing Examples 428 Handling Lists of Lists Using the SequenceExample Protobuf 429 Preprocessing the Input Features 430 Encoding Categorical Features Using One-Hot Vectors 431 Encoding Categorical Features Using Embeddings 433 Keras Preprocessing Layers 437 TF Transform 439 The TensorFlow Datasets (TFDS) Project 441 Exercises 442 14. Deep Computer Vision Using Convolutional Neural Networks. . . . . . . . . . . . . . . . . . . 445 The Architecture of the Visual Cortex 446 Convolutional Layers 448 Filters 450 Stacking Multiple Feature Maps 451 Table of Contents | ix TensorFlow Implementation 453 Memory Requirements 456 Pooling Layers 456 TensorFlow Implementation 458 CNN Architectures 460 LeNet-5 463 AlexNet 464 GoogLeNet 466 VGGNet 470 ResNet 471 Xception 474 SENet 476 Implementing a ResNet-34 CNN Using Keras 478 Using Pretrained Models from Keras 479 Pretrained Models for Transfer Learning 481 Classification and Localization 483 Object Detection 485 Fully Convolutional Networks 487 You Only Look Once (YOLO) 489 Semantic Segmentation 492 Exercises 496 15. Processing Sequences Using RNNs and CNNs. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 497 Recurrent Neurons and Layers 498 Memory Cells 500 Input and Output Sequences 501 Training RNNs 502 Forecasting a Time Series 503 Baseline Metrics 505 Implementing a Simple RNN 505 Deep RNNs 506 Forecasting Several Time Steps Ahead 508 Handling Long Sequences 511 Fighting the Unstable Gradients Problem 512 Tackling the Short-Term Memory Problem 514 Exercises 523 16. Natural Language Processing with RNNs and Attention. . . . . . . . . . . . . . . . . . . . . . . . 525 Generating Shakespearean Text Using a Character RNN 526 Creating the Training Dataset 527 How to Split a Sequential Dataset 527 Chopping the Sequential Dataset into Multiple Windows 528 x | Table of Contents Building and Training the Char-RNN Model 530 Using the Char-RNN Model 531 Generating Fake Shakespearean Text 531 Stateful RNN 532 Sentiment Analysis 534 Masking 538 Reusing Pretrained Embeddings 540 An Encoder–Decoder Network for Neural Machine Translation 542 Bidirectional RNNs 546 Beam Search 547 Attention Mechanisms 549 Visual Attention 552 Attention Is All You Need: The Transformer Architecture 554 Recent Innovations in Language Models 563 Exercises 565 17. Representation Learning and Generative Learning Using Autoencoders and GANs. 567 Efficient Data Representations 569 Performing PCA with an Undercomplete Linear Autoencoder 570 Stacked Autoencoders 572 Implementing a Stacked Autoencoder Using Keras 572 Visualizing the Reconstructions 574 Visualizing the Fashion MNIST Dataset 574 Unsupervised Pretraining Using Stacked Autoencoders 576 Tying Weights 577 Training One Autoencoder at a Time 578 Convolutional Autoencoders 579 Recurrent Autoencoders 580 Denoising Autoencoders 581 Sparse Autoencoders 582 Variational Autoencoders 586 Generating Fashion MNIST Images 590 Generative Adversarial Networks 592 The Difficulties of Training GANs 596 Deep Convolutional GANs 598 Progressive Growing of GANs 601 StyleGANs 604 Exercises 607 18. Reinforcement Learning. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 609 Learning to Optimize Rewards 610 Policy Search 612 Table of Contents | xi Introduction to OpenAI Gym 613 Neural Network Policies 617 Evaluating Actions: The Credit Assignment Problem 619 Policy Gradients 620 Markov Decision Processes 625 Temporal Difference Learning 629 Q-Learning 630 Exploration Policies 632 Approximate Q-Learning and Deep Q-Learning 633 Implementing Deep Q-Learning 634 Deep Q-Learning Variants 639 Fixed Q-Value Targets 639 Double DQN 640 Prioritized Experience Replay 640 Dueling DQN 641 The TF-Agents Library 642 Installing TF-Agents 643 TF-Agents Environments 643 Environment Specifications 644 Environment Wrappers and Atari Preprocessing 645 Training Architecture 649 Creating the Deep Q-Network 650 Creating the DQN Agent 652 Creating the Replay Buffer and the Corresponding Observer 654 Creating Training Metrics 655 Creating the Collect Driver 656 Creating the Dataset 658 Creating the Training Loop 661 Overview of Some Popular RL Algorithms 662 Exercises 664 19. Training and Deploying TensorFlow Models at Scale. . . . . . . . . . . . . . . . . . . . . . . . . . . 667 Serving a TensorFlow Model 668 Using TensorFlow Serving 668 Creating a Prediction Service on GCP AI Platform 677 Using the Prediction Service 682 Deploying a Model to a Mobile or Embedded Device 685 Using GPUs to Speed Up Computations 689 Getting Your Own GPU 690 Using a GPU-Equipped Virtual Machine 692 Colaboratory 693 Managing the GPU RAM 694 xii | Table of Contents Placing Operations and Variables on Devices 697 Parallel Execution Across Multiple Devices 699 Training Models Across Multiple Devices 701 Model Parallelism 701 Data Parallelism 704 Training at Scale Using the Distribution Strategies API 709 Training a Model on a TensorFlow Cluster 711 Running Large Training Jobs on Google Cloud AI Platform 714 Black Box Hyperparameter Tuning on AI Platform 716 Exercises 717 Thank You! 718 A. Exercise Solutions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 719 B. Machine Learning Project Checklist. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 755 C. SVM Dual Problem. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 761 D. Autodiff. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 765 E. Other Popular ANN Architectures. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 773 F. Special Data Structures. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 783 G. TensorFlow Graphs. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 791 Index. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 801 9789355421982

]]> Place hold on Hands-on machine learning with scikit-learn, keras, and tensorflow

Data visualization with Python and JavaScript : scrape, clean, explore, and transform your data

]]> By Dale, Kyran .
New Delhi Shroff publishers & distributors 2023 .
xxxvi, 529 , Includes index 9789355422392

]]> Place hold on Data visualization with Python and JavaScript