Book Image

AI Blueprints

By : Dr. Joshua Eckroth, Eric Schoen
Book Image

AI Blueprints

By: Dr. Joshua Eckroth, Eric Schoen

Overview of this book

AI Blueprints gives you a working framework and the techniques to build your own successful AI business applications. You’ll learn across six business scenarios how AI can solve critical challenges with state-of-the-art AI software libraries and a well thought out workflow. Along the way you’ll discover the practical techniques to build AI business applications from first design to full coding and deployment. The AI blueprints in this book solve key business scenarios. The first blueprint uses AI to find solutions for building plans for cloud computing that are on-time and under budget. The second blueprint involves an AI system that continuously monitors social media to gauge public feeling about a topic of interest - such as self-driving cars. You’ll learn how to approach AI business problems and apply blueprints that can ensure success. The next AI scenario shows you how to approach the problem of creating a recommendation engine and monitoring how those recommendations perform. The fourth blueprint shows you how to use deep learning to find your business logo in social media photos and assess how people interact with your products. Learn the practical techniques involved and how to apply these blueprints intelligently. The fifth blueprint is about how to best design a ‘trending now’ section on your website, much like the one we know from Twitter. The sixth blueprint shows how to create helpful chatbots so that an AI system can understand customers’ questions and answer them with relevant responses. This book continuously demonstrates a working framework and strategy for building AI business applications. Along the way, you’ll also learn how to prepare for future advances in AI. You’ll gain a workflow and a toolbox of patterns and techniques so that you can create your own smart code.
Table of Contents (14 chapters)
AI Blueprints
Foreword
Contributors
Preface
Other Books You May Enjoy
Index

Recognizing anomalies


Anomalies are a different but related kind of information from trends. While trend analysis aims to discover what is normal about a data stream, recognizing anomalies is about finding out which events represented in the data stream are clearly abnormal. To recognize anomalies, one must already have an idea of what is normal. Additionally, recognizing anomalies requires deciding some threshold of how far from normal data may be before it is labeled anomalous.

We will look at four techniques for recognizing anomalies. First, we'll devise two ways to use z-scores to identify data points that are significantly different from the average data point. Then we will look at a variation of principal component analysis, a kind of matrix decomposition technique similar to singular value decomposition from Chapter 4, A Blueprint for Recommending Products and Services, that separates normal data from anomalous or extreme events from noise. Finally, we will use a cosine similarity...