Applied Machine Learning and AI for Engineers: Solve Business Problems That Can’t Be Solved Algorithmically

https://kanchisilksarees.com/2a09vtj Foreword Preface Who Should Read This Book Why I Wrote This Book Running the Book’s Code Samples Navigating This Book Conventions Used in This Book Using Code Examples O’Reilly Online Learning How to Contact Us Acknowledgments I. Machine Learning with Scikit-Learn 1. Machine Learning What Is Machine Learning? Machine Learning Versus Artificial Intelligence Supervised Versus Unsupervised Learning Unsupervised Learning with k-Means Clustering Applying k-Means Clustering to Customer Data Segmenting Customers Using More Than Two Dimensions Supervised Learning k-Nearest Neighbors Using k-Nearest Neighbors to Classify Flowers Summary 2. Regression Models Linear Regression Decision Trees Random Forests Gradient-Boosting Machines Support Vector Machines Accuracy Measures for Regression Models Using Regression to Predict Taxi Fares Summary 3. Classification Models Logistic Regression Accuracy Measures for Classification Models Categorical Data Binary Classification Classifying Passengers Who Sailed on the Titanic Detecting Credit Card Fraud Multiclass Classification Building a Digit Recognition Model Summary 4. Text Classification Preparing Text for Classification Sentiment Analysis Naive Bayes Spam Filtering Recommender Systems Cosine Similarity Building a Movie Recommendation System Summary 5. Support Vector Machines How Support Vector Machines Work Kernels Kernel Tricks Hyperparameter Tuning Data Normalization Pipelining Using SVMs for Facial Recognition Summary 6. Principal Component Analysis Understanding Principal Component Analysis Filtering Noise Anonymizing Data Visualizing High-Dimensional Data Anomaly Detection Using PCA to Detect Credit Card Fraud Using PCA to Predict Bearing Failure Multivariate Anomaly Detection Summary 7. Operationalizing Machine Learning Models Consuming a Python Model from a Python Client Versioning Pickle Files Consuming a Python Model from a C# Client Containerizing a Machine Learning Model Using ONNX to Bridge the Language Gap Building ML Models in C# with ML.NET Sentiment Analysis with ML.NET Saving and Loading ML.NET Models Adding Machine Learning Capabilities to Excel Summary II. Deep Learning with Keras and TensorFlow 8. Deep Learning Understanding Neural Networks Training Neural Networks Summary 9. Neural Networks Building Neural Networks with Keras and TensorFlow Sizing a Neural Network Using a Neural Network to Predict Taxi Fares Binary Classification with Neural Networks Making Predictions Training a Neural Network to Detect Credit Card Fraud Multiclass Classification with Neural Networks Training a Neural Network to Recognize Faces Dropout Saving and Loading Models Keras Callbacks Summary 10. Image Classification with Convolutional Neural Networks Understanding CNNs Using Keras and TensorFlow to Build CNNs Training a CNN to Recognize Arctic Wildlife Pretrained CNNs Using ResNet50V2 to Classify Images Transfer Learning Using Transfer Learning to Identify Arctic Wildlife Data Augmentation Image Augmentation with ImageDataGenerator Image Augmentation with Augmentation Layers Applying Image Augmentation to Arctic Wildlife Global Pooling Audio Classification with CNNs Summary 11. Face Detection and Recognition Face Detection Face Detection with Viola-Jones Using the OpenCV Implementation of Viola-Jones Face Detection with Convolutional Neural Networks Extracting Faces from Photos Facial Recognition Applying Transfer Learning to Facial Recognition Boosting Transfer Learning with Task-Specific Weights ArcFace Putting It All Together: Detecting and Recognizing Faces in Photos Handling Unknown Faces: Closed-Set Versus Open-Set Classification Summary 12. Object Detection R-CNNs Mask R-CNN YOLO YOLOv3 and Keras Custom Object Detection Training a Custom Object Detection Model with the Custom Vision Service Using the Exported Model Summary 13. Natural Language Processing Text Preparation Word Embeddings Text Classification Automating Text Vectorization Using TextVectorization in a Sentiment Analysis Model Factoring Word Order into Predictions Recurrent Neural Networks (RNNs) Using Pretrained Models to Classify Text Neural Machine Translation LSTM Encoder-Decoders Transformer Encoder-Decoders Building a Transformer-Based NMT Model Using Pretrained Models to Translate Text Bidirectional Encoder Representations from Transformers (BERT) Building a BERT-Based Question Answering System Fine-Tuning BERT to Perform Sentiment Analysis Summary 14. Azure Cognitive Services Introducing Azure Cognitive Services Keys and Endpoints Calling Azure Cognitive Services APIs Azure Cognitive Services Containers The Computer Vision Service The Language Service The Translator Service The Speech Service Putting It All Together: Contoso Travel Summary Index