Machine Learning by Tutorials (Second Edition): Beginning Machine Learning for Apple and iOS

Learn Machine Learning! Machine learning is one of those topics that can be daunting at first blush. It's not clear where to start, what path someone should take and what APIs to learn in order to get started teaching machines how to learn. This is where Machine Learning by Tutorials comes in! In this book, we'll hold your hand through a number of tutorials, to get you started in the world of machine learning. We'll cover a wide range of popular topics in the field of machine learning, while developing apps that work on iOS devices. Who This Book Is For This books is for the intermediate iOS developer who already knows the basics of iOS and Swift development, but wants to understand how machine learning works. Topics covered in Machine Learning by Tutorials CoreML: Learn how to add a machine learning model to your iOS apps, and how to use iOS APIs to access it. Create ML: Learn how to create your own model using Apple's Create ML Tool. Turi Create and Keras: Learn how to tune parameters to improve your machine learning model using more advanced tools. Image Classification: Learn how to apply machine learning models to predict objects in an image. Convolutional Networks: Learn advanced machine learning techniques for predicting objects in an image with Convolutional Neural Networks (CNNs). Sequence Classification: Learn how you can use recurrent neural networks (RNNs) to classify motion from an iPhone's motion sensor. Text-to-text Transform: Learn how to use machine learning to convert bodies of text between two languages. By the end of this book, you'll have a firm understanding of what machine learning is, what it can and cannot do, and how you can use machine learning in your next app! About the Cover What You Need Book License Book Source Code & Forums Chapter 1: Machine Learning, iOS & You What is machine learning? Deep learning ML in a nutshell Can mobile devices really do machine learning? Frameworks, tools and APIs ML all the things? The ethics of machine learning Key points Where to go from here? Chapter 2: Getting Started with Image Classification Is that snack healthy? Core ML Vision Creating the VNCoreML request Performing the request Showing the results How does it work? Multi-class classification Bonus: Using Core ML without Vision Challenge Key points Chapter 3: Training the Image Classifier The dataset Create ML How we created the dataset Transfer learning Logistic regression Looking for validation More metrics and the test set Examining Your Output Model Recap Challenge Key points Chapter 4: Getting Started with Python & Turi Create Starter folder Python Packages and environments Installing Anaconda Useful Conda commands Setting up a base ML environment Jupyter Notebooks Transfer learning with Turi Create Shutting down Jupyter Docker and Colab Key points Where to go from here? Chapter 5: Digging Deeper into Turi Create Getting started Transfer learning with SqueezeNet Getting individual predictions Using a fixed validation set Increasing max iterations Confusing apples with oranges? Training the classifier with regularization Wrangling Turi Create code A peek behind the curtain Challenges Key points Chapter 6: Taking Control of Training with Keras Getting started Back to basics with logistic regression Building the model Loading the data Training the logistic regression model Your first neural network Challenge Key points Chapter 7: Going Convolutional Got GPU? Convolution layers Your first convnet in Keras Key points Where to go from here? Chapter 8: Advanced Convolutional Neural Networks SqueezeNet MobileNet and data augmentation How good is the model really? Converting to Core ML Challenges Key points Chapter 9: Beyond Classification Where is it? A simple localization model Key points Chapter 10: YOLO & Semantic Segmentation Single stage detectors Hello Turi, my old friend The demo app Semantic segmentation Challenges Key points Where to go from here? Chapter 11: Data Collection for Sequence Classification Building a dataset Analyzing and preparing your data Key points Where to go from here? Chapter 12: Training a Model for Sequence Classification Creating a model Getting to know your model A note on sequence classification Key points Where to go from here? Chapter 13: Sequence Classification Classifying human activity in your app Challenges Key points Chapter 14: Natural Language Classification Getting started Language identification Finding named entities Adding a search feature Sentiment analysis Building a sentiment classifier Custom word classifiers The remaining bits Key points Where to go from here? Chapter 15: Natural Language Transformation, Part 1 Getting started The sequence-to-sequence model Prepare your dataset Build your model Train your model Inference with sequence-to-sequence models Converting your model to Core ML Using your model in iOS Let's talk translation quality Key points Where to go from here? Chapter 16: Natural Language Transformation, Part 2 Bidirectional RNNs Beam search Attention Why use characters at all? Words as tokens and word embedding Key points Where to go from here? Conclusion Photo Credits