Chapter 1, Introduction to Machine Learning and Predictive Analytics, this introductory chapter is a general presentation of the Amazon Machine Learning service and the types of predictive analytics problems it can address. We show how the service uses a simple linear model for regression and classification problems, and we present the context for successful predictions with Amazon Machine Learning.

Chapter 2, Machine Learning Definitions and Concepts, this chapter explains the machine learning concepts needed to use the Amazon Machine Learning service and fully understand how it works. What are the preparation techniques used when dealing with raw data? How do we evaluate a predictive model? What strategies are available to remediate poor predictive performances?

Chapter 3, Overview of an Amazon Machine Learning Workflow, this chapter is an overview of a simple Amazon Machine Learning project. The reader will learn how to get started on the Amazon Machine Learning platform, how to set up an account, and how to secure it. We go through a simple numeric prediction problem based on a classic dataset. We describe how to prepare the data, train and select a model, and make predictions.

Chapter 4, Loading and Preparing the Dataset, Amazon ML offers powerful features to transform the data through recipes. Working with a classic dataset, we upload data on S3, implement cross validation, create a schema, and examine available data statistics. We then extend Amazon ML feature engineering and data cleaning capabilities by using Athena, a recently launched AWS SQL service.

Chapter 5, Model Creation, in this chapter, we explore the Amazon ML data transformations and how to apply them through recipes. We train and tune models and select the best ones by analyzing different prediction metrics. We present insights into the Stochastic Gradient Descent algorithm, and the use of different types of regularization. Finally, we analyze the training logs to better understand what goes on under the Amazon ML hood during model training.

Chapter 6,Predictions and Performances, we apply our newly trained models to make predictions on previously unseen data, and we make a final assessment of their performance and robustness. We show how to make batch predictions on a given dataset and how to set up a real-time endpoint for streaming predictions.

Chapter 7, Command Line and SDK, using the AWS web interface to manage and run your projects is time-consuming. In this chapter, we move away from the web interface and start running our projects via the command line with the AWS Command Line Interface (AWS CLI) and the Python SDK with the Boto3 library. We use our new powers to implement cross validation and recursive feature selection.

Chapter 8, Creating Datasources from Redshift, in this chapter, we will use the power of SQL queries to address non-linear datasets. Creating datasources in Redshift gives us the potential for upstream SQL based feature engineering prior to datasource creation. We explore how to upload data from S3 to Redshift, access the database, run queries, and export results.

Chapter 9, Building a Streaming Data Analysis Pipeline, in the final chapter of the book, we extend Amazon ML capabilities by integrating with other AWS services. We build a fully featured streaming data flow integrating AWS Kinesis, Lambda, Redshift, and Machine Learning to implement real time tweets classification.

This book requires an AWS account, and the ability to run Python 3 (3.5) code examples. We use the latest Anaconda distribution (conda 4.3). The reader should also be able to run basic shell commands in a terminal.  Note that the Amazon Machine Learning service is not part of the free tier AWS offer.

This book is for Python developers who want to build real-world artificial intelligence applications. This book is friendly to Python beginners, but being familiar with Python would be useful to play around with the code. It will also be useful for experienced Python programmers who are looking to use artificial intelligence techniques in their existing technology stacks.

In this book, you will find a number of text styles that distinguish between different kinds of information. Here are some examples of these styles and an explanation of their meaning.

Code words in text, database table names, folder names, filenames, file extensions, pathnames, dummy URLs, user input, and Twitter handles are shown as follows: "Set original weight estimation at w_0 = 100g to initialize and a counter."

A block of code is set as follows:

# Create datasource for training
resource = name_id_generation('DS', 'training', trial)
print("Creating datasources for training (%s)"% resource['Name'] )

Any command-line input or output is written as follows:

$ aws s3 cp data/titanic.csv s3://aml.packt/data/ch9/

New terms and important words are shown in bold. Words that you see on the screen, for example, in menus or dialog boxes, appear in the text like this: "Examples of reinforcement learning applications include AlphaGo, Google's world championship Go algorithm, self-driving cars, and semi-autonomous robots."

Feedback from our readers is always welcome. Let us know what you think about this book-what you liked or disliked. Reader feedback is important for us as it helps us develop titles that you will really get the most out of. To send us general feedback, simply e-mail [email protected], and mention the book's title in the subject of your message. If there is a topic that you have expertise in and you are interested in either writing or contributing to a book, see our author guide at www.packtpub.com/authors.

Now that you are the proud owner of a Packt book, we have a number of things to help you to get the most from your purchase.