Chapter 1, Introduction to IBM Cloud, provides a brief introduction to the IBM cloud platform and the machine learning service. Moreover, this chapter provides detailed instructions on how to set up data science and machine learning development environments on IBM Cloud. Finally, it conclude in with an example for loading and visualizing data.
Chapter 2, Feature Extraction – A Bag of Tricks, provides a hands-on guide to extraction and the selection of features from real-life data, with an emphasis on the fact that practical machine learning systems are all about proper feature engineering. This chapter demonstrates best practices for feeding data to your machine learning algorithms. Moreover, it shows how to remove redundant data that negatively impacts the performance of your machine learning system. This chapter also introduces strategies for combining data from different sources.
Chapter 3, Supervised Machine Learning Models for Your Data, acts as the backbone of the entire book. It provides a tour of the machine learning paradigm, with a focus on famous approaches and algorithms. This chapter starts by providing a practical background to model evaluation, model selection, and algorithm selection in machine learning. Then it covers supervised learning, and discusses machine learning algorithms for regression problems. By the end of the chapter, you should be able to select proper supervised machine learning models for the data at hand.
Chapter 4, Implementing Unsupervised Algorithms, is the sequel to the tour of the machine learning paradigm that began in Chapter 2, Feature Extraction – A Bag of Tricks. This chapter covers unsupervised learning and semi-supervised learning. Moreover, it discusses famous clustering algorithms, before concluding with a discussion of online versus batch learning.
Chapter 5, Machine Learning Workouts on IBM Cloud, provides several examples that are carefully designed to uncover the power of Python as a machine learning programming language of choice and the power of the machine learning service on the IBM Watson Studio platform. This chapter will enable you to practice proper feature engineering. Moreover, you will be able to run supervised (classification) and unsupervised (clustering) techniques in the cloud. Furthermore, this chapter will guide you on implementing time series prediction algorithms. Finally, it concludes with visual recognition examples.
Chapter 6, Using Spark with IBM Watson Studio, provides guidelines for creating a Spark machine learning pipeline within IBM Watson Studio.
Chapter 7, Deep Learning Using TensorFlow on IBM Cloud, provides an introduction to deep learning and neural networks on IBM Cloud. An overview of the use of TensorFlow on the cloud will also be provided. This chapter is designed to balance theory and practical implementation.
Chapter 8, Creating a Facial Expression Platform on IBM Cloud, covers a complete, cloud-based facial expression classification solution using deep learning. It implements a simple, yet efficient, neural network model using TensorFlow and the machine learning service on Watson Studio. This chapter demonstrates an end-to-end solution for a complex machine learning task.
Chapter 9, The Automated Classification of Lithofacies Formation Using ML, demonstrates a cloud-based machine learning system for identifying lithofacies based on well-log measurements. This is a crucial step in drilling applications. First, we will begin by introducing the problem and the dataset. Then we will explain the types of post-processing needed for such cases. Finally, the complete solution is built using the machine learning service on the Watson platform.
Chapter 10, Building a Cloud-Based Multibiometric Identity Authentication Platform, guides the reader on how to build a complete cloud-based human identification system using biometric traits. This chapter begins by introducing the problem and the datasets. Then it explains the types of post-processing needed for each biometric. Moreover, you will learn how to extract meaningful features. Learning and recognition will follow. Finally, you will practice multimodal data fusion.
Chapter 11, Conclusion, concludes the book with an overview of what has been covered. This chapter also sheds some light on some of the practical considerations related to developing machine learning systems on the cloud. Finally, the book concludes with further discussions on different routes the reader might consider to enhance their skills.
